File size: 73,101 Bytes
6fa4bc9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
{
    "paper_id": "2020",
    "header": {
        "generated_with": "S2ORC 1.0.0",
        "date_generated": "2023-01-19T07:28:59.508073Z"
    },
    "title": "Arabic NLG Language Functions",
    "authors": [
        {
            "first": "Wael",
            "middle": [],
            "last": "Abed",
            "suffix": "",
            "affiliation": {
                "laboratory": "",
                "institution": "University of Aberdeen",
                "location": {
                    "postCode": "AB24 3UE",
                    "settlement": "Aberdeen",
                    "country": "UK"
                }
            },
            "email": ""
        },
        {
            "first": "Ehud",
            "middle": [],
            "last": "Reiter",
            "suffix": "",
            "affiliation": {
                "laboratory": "",
                "institution": "University of Aberdeen",
                "location": {
                    "postCode": "AB24 3UE",
                    "settlement": "Aberdeen",
                    "country": "UK"
                }
            },
            "email": "ereiter@csd.abdn.ac.uk"
        }
    ],
    "year": "",
    "venue": null,
    "identifiers": {},
    "abstract": "The Arabic language has very limited supports from NLG researchers. In this paper, we explain the challenges of the core grammar, provide a lexical resource, and implement the first language functions for the Arabic language. We did a human evaluation to evaluate our functions in generating sentences from the NADA Corpus.",
    "pdf_parse": {
        "paper_id": "2020",
        "_pdf_hash": "",
        "abstract": [
            {
                "text": "The Arabic language has very limited supports from NLG researchers. In this paper, we explain the challenges of the core grammar, provide a lexical resource, and implement the first language functions for the Arabic language. We did a human evaluation to evaluate our functions in generating sentences from the NADA Corpus.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Abstract",
                "sec_num": null
            }
        ],
        "body_text": [
            {
                "text": "The Arabic Language is a member of the Semitic language family, the fifth most spoken language in the world. Around 330 million people speak this language as their mother tongue language distributed in 25 countries. It is mainly classified into two types; Classical and Modern Standard Arabic (MSA) (Al-Radaideh, 2020) , in addition to regional variances (dialects). As far as we know, there are not any pipeline NLG systems or libraries that support an Arabic surface realiser at the time of writing this paper. Two factors might be the reasons: Arabic Language complexity and limited resources. An example that reflects the second cause is that even the Google n-gram viewer does not support the Arabic language (Alsmadi and Zarour, 2018) . Therefore, in this paper, we are presenting six simple morphological MSA natural language functions (inflectNoun, buildNP, inflectVerb, inflectAdjective, pronouns and numToWords). These functions can be considered as a good start to build an Arabic surface realiser, which can be used in a pipeline NLG system as they can be used to correctly generate and inflect different parts of speech including the generation of a noun phrase.",
                "cite_spans": [
                    {
                        "start": 299,
                        "end": 318,
                        "text": "(Al-Radaideh, 2020)",
                        "ref_id": "BIBREF5"
                    },
                    {
                        "start": 714,
                        "end": 740,
                        "text": "(Alsmadi and Zarour, 2018)",
                        "ref_id": "BIBREF7"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "We implemented these functions from scratch using a rule-based approach. Furthermore, a lexicon containing around 9000 lemmata were collected. An open-source versions of lexicon and code with a web interface are provided and available online on GitHub. 1 .",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "As there is no NLG system done for Arabic, we investigated similar approaches that targets the language challenges. Green and DeNero (2012) presented a method to cover words agreements to generate correct words inflections in a translating task using a class-based agreement model. While, Khoufi et al. (2016) dealt with un-vocalisation and agglutination in Arabic. They built a statistical model to extract grammatical rules from a treebank corpus to use these rules for parsing.",
                "cite_spans": [
                    {
                        "start": 116,
                        "end": 139,
                        "text": "Green and DeNero (2012)",
                        "ref_id": null
                    },
                    {
                        "start": 282,
                        "end": 309,
                        "text": "While, Khoufi et al. (2016)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Related Works",
                "sec_num": "2"
            },
            {
                "text": "Furthermore, another member of the Semitic group, which is the Hebrew language, also lacks NLG resources. One attempt to Generate a noun compounds in Hebrew goes back to 1998 (Netzer and Elhadad, 1998) , they used a rule-based approach and listed the difficulties between English and Hebrew in terms of the syntactic, semantic and lexical constraints for generating a Noun Compounds, including a special feature of Hebrew called \"Smixut\".",
                "cite_spans": [
                    {
                        "start": 175,
                        "end": 201,
                        "text": "(Netzer and Elhadad, 1998)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Related Works",
                "sec_num": "2"
            },
            {
                "text": "For other languages, researchers used different realisers' technologies such as handcrafted grammars rule-based (Gatt and Reiter, 2009) , statistical approaches (Gardent and Perez-Beltrachini, 2017) and human-generated template-based (Gatt and Krahmer, 2018) . However, SimpleNLG is still considered the most popular method since it was presented by Gatt and Reiter (2009) , It is a rulebased realiser that uses three main steps syntactic, morphological and orthographic realisations to produce a correctly inflected and high-quality output. It was adapted to many languages includ-ing (in chronological order) French (Vaudry and Lapalme, 2013), Italian (Mazzei et al., 2016) , Spanish (Ramos-Soto et al., 2017) , Galician (Cascallar-Fuentes et al., 2018) , Dutch (de Jong and Theune, 2018), Mandarin (Chen et al., 2018) and German (Braun et al., 2019) .",
                "cite_spans": [
                    {
                        "start": 112,
                        "end": 135,
                        "text": "(Gatt and Reiter, 2009)",
                        "ref_id": null
                    },
                    {
                        "start": 161,
                        "end": 198,
                        "text": "(Gardent and Perez-Beltrachini, 2017)",
                        "ref_id": "BIBREF12"
                    },
                    {
                        "start": 234,
                        "end": 258,
                        "text": "(Gatt and Krahmer, 2018)",
                        "ref_id": null
                    },
                    {
                        "start": 350,
                        "end": 372,
                        "text": "Gatt and Reiter (2009)",
                        "ref_id": null
                    },
                    {
                        "start": 646,
                        "end": 675,
                        "text": "Italian (Mazzei et al., 2016)",
                        "ref_id": null
                    },
                    {
                        "start": 678,
                        "end": 711,
                        "text": "Spanish (Ramos-Soto et al., 2017)",
                        "ref_id": null
                    },
                    {
                        "start": 723,
                        "end": 755,
                        "text": "(Cascallar-Fuentes et al., 2018)",
                        "ref_id": "BIBREF9"
                    },
                    {
                        "start": 801,
                        "end": 820,
                        "text": "(Chen et al., 2018)",
                        "ref_id": "BIBREF10"
                    },
                    {
                        "start": 832,
                        "end": 852,
                        "text": "(Braun et al., 2019)",
                        "ref_id": "BIBREF8"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Related Works",
                "sec_num": "2"
            },
            {
                "text": "Another popular approach is to provide grammars' functions to be used in a template-based NLG system. Such as RosaeNLG 2 ,which is an open-source, template-based, NLG library that gives simpler functions than SimpleNLG and had been used for English, French, German, Italian and Spanish languages. Similar approaches with fewer options can be found in Core-NLG 3 and xSpin 4 .",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Related Works",
                "sec_num": "2"
            },
            {
                "text": "Our approach is more similar to RosaeNLG than to SimpleNLG. As we will provide language functions and a noun phrase function to give correct grammatical inflections and agreements. Due to the huge morphological difference between the Semitic and non-Semitic languages, we could not adopt SimpleNLG nor RosaeNLG. Therefore, we found it convenient as a start to building our functions from scratch following the same approach of RosaeNLG.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Related Works",
                "sec_num": "2"
            },
            {
                "text": "The Arabic language consists of 28 alphabet letters, written from right to left and there is no letters capitalisation. It has a very rich morphology measured in the number of different forms for its nouns, verbs and particles (Al-Radaideh, 2020) .",
                "cite_spans": [
                    {
                        "start": 227,
                        "end": 246,
                        "text": "(Al-Radaideh, 2020)",
                        "ref_id": "BIBREF5"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Arabic language challenges",
                "sec_num": "3"
            },
            {
                "text": "Arabic nouns can be inflected depending on the sentence gender, number, and grammatical case (nominative, genitive, and accusative).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Nouns",
                "sec_num": "3.1"
            },
            {
                "text": "Arabic assigns gender for living and non-living things (ex: \"car\" ( ) treated as feminine, while \"Goal\" ( ) is a masculine noun) (Shamsan and Attayib, 2015). Moreover, unlike many languages, Arabic has a dual form in addition to singular and plural form. This form is used to express a noun that refers to two entities (things or people), and can be generated by adding one of these suffixes (\" \"-aan, \" \"-ayn, \" \"-taan, \" \"-tayn) to the singular form). The type of suffix used depends Add prefix \"ah\" ( ) and infix (w) ( ) Note: in the examples above there are other changes in the small vowels marks that did not mention because it will not be covered here Regarding the plural form, there are two different types: sound plural and broken plural. The first one can be inflected using similar steps to the steps used in dual form, with a different set of suffixes (\" \"uun, \" \"-iin, \" \"-aat). Some cases can only be identified by the \"h . arak\u00e3t\" ( ) short vowel marks of the last consonant letter. The short vowel marks will be neglected here, as for MSA according to many studies, adding these \"Short Vowels Do Not Contribute to the Quality of Reading\" (Abu-Rabia, 2019).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Nouns",
                "sec_num": "3.1"
            },
            {
                "text": "While the Broken plural has different inflections. This type has more than 30 patterns to use with internal and external manipulation of the sounds and vowels letters in the word and there are no firm rules (Dawdy-Hesterberg and Pierrehumbert, 2014). Table 1 illustrates all these states for dual, sound plural and broken plural.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 251,
                        "end": 258,
                        "text": "Table 1",
                        "ref_id": "TABREF1"
                    }
                ],
                "eq_spans": [],
                "section": "Nouns",
                "sec_num": "3.1"
            },
            {
                "text": "The challenges of the Arabic numerals are their relation with the noun's gender, case, definiteness, word order and counts. In addition, they do not follow the rules in some cases (Al-Bataineh and Branigan, 2020) . These facts classify Arabic numerals as a complex problem compared to its relevance in the English language. Arabic numerals (ordinal and cardinal) can be classified mainly into three groups (Al-Bataineh and Branigan, 2020) Simplex numerals (1-10): according to (Al Easa, 1997) this group has two subgroups: singular numerals (1 and 2) and added numerals (3-10). Singular numerals agree with the counted noun in all aspects. However, this is the only group in the Arabic numerals that comes after the noun in the word order. For the second subgroup, the numeral gender will take the opposite form of the noun gender (see table 2).",
                "cite_spans": [
                    {
                        "start": 180,
                        "end": 212,
                        "text": "(Al-Bataineh and Branigan, 2020)",
                        "ref_id": "BIBREF3"
                    },
                    {
                        "start": 406,
                        "end": 438,
                        "text": "(Al-Bataineh and Branigan, 2020)",
                        "ref_id": "BIBREF3"
                    },
                    {
                        "start": 477,
                        "end": 492,
                        "text": "(Al Easa, 1997)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Numerals",
                "sec_num": "3.2"
            },
            {
                "text": "Compound numerals (11-19): It has two major differences. The first one is that each numeral consists of two words, one for each digit (units and ten). For example, to say \"twelve\" in Arabic it will be \"itn\u00e3 'asara\" ( ) which means \"two\" \"ten\". The second is the gender agreement. For \"eleven\" and \"twelve\", both parts of the numerals agree with the noun (Al Easa, 1997) . While the rest from \"thirteen\" to \"nineteen\", the first part takes the inverse gender of the noun and the second part agrees (Al-Bataineh and Branigan, 2020) .",
                "cite_spans": [
                    {
                        "start": 354,
                        "end": 369,
                        "text": "(Al Easa, 1997)",
                        "ref_id": null
                    },
                    {
                        "start": 497,
                        "end": 529,
                        "text": "(Al-Bataineh and Branigan, 2020)",
                        "ref_id": "BIBREF3"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Numerals",
                "sec_num": "3.2"
            },
            {
                "text": "Complex or doubled numerals (20-): One word is used for each digit and they connected by the conjunction \"and\" (\" \" pronounced \"wa\"). The first digit (units) will follow the \"Simplex numerals\" rules (except for the word order). While the \"tens\" have one form which is the masculine plural form. The third digit \"hundred\" (\" \" or \" \" pronounced as \"mi'a\") considered as a feminine noun. The rest (fourth, fifth . . . etc.) digits treated as masculine nouns and use their plural form with numerals greater than two.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Numerals",
                "sec_num": "3.2"
            },
            {
                "text": "For ordinal numbers, the rules are mostly similar. However, there are a few differences. Such as ordinal numbers always agree with the noun gender, and nouns come before the numerals (Abdel-Malek et al., 1983) .",
                "cite_spans": [
                    {
                        "start": 183,
                        "end": 209,
                        "text": "(Abdel-Malek et al., 1983)",
                        "ref_id": "BIBREF0"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Numerals",
                "sec_num": "3.2"
            },
            {
                "text": "In MSA, Adjectives commonly used after the nouns, although they can be used before as well. They agree with the noun in gender, case, definiteness and number (Shamsan and Attayib, 2015) (Idrissi et al., 2019) . If the adjective is used to describe an un-human plural noun, then the adjective will take the singular feminine form (Shamsan and Attayib, 2015). However, with time this exception started to disappear and full agreement in the nounadjective sentence is used (Idrissi et al., 2019) . Ex:",
                "cite_spans": [
                    {
                        "start": 186,
                        "end": 208,
                        "text": "(Idrissi et al., 2019)",
                        "ref_id": null
                    },
                    {
                        "start": 470,
                        "end": 492,
                        "text": "(Idrissi et al., 2019)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Adjectives",
                "sec_num": "3.3"
            },
            {
                "text": "-the-men (M-pluraldefinite-nominative) the-wise (M-plural-definitenominative)-The wise men",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Adjectives",
                "sec_num": "3.3"
            },
            {
                "text": "Arabic verbs participate in a large percentage of the sentence due to their morphological richness. There are 19 forms for verb inflections, and for each form, the verb agrees with the subject in terms of number, person and gender. Moreover, it can be inflected using four moods, two voices and three tenses (Shaalan et al., 2015).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Verbs",
                "sec_num": "3.4"
            },
            {
                "text": "Arabic pronouns can be categorised to Explicit and Implicit forms. The latter will not be discussed here as it only occurs in spoken Arabic. On the other hand, there are 3 branches can be derived from the Explicit form: attached, detached and reflexive (Igaab and Tarrad, 2019). Which can be used to create personal, demonstrative and relative pronouns (Ryding, 2005) .",
                "cite_spans": [
                    {
                        "start": 353,
                        "end": 367,
                        "text": "(Ryding, 2005)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Pronouns",
                "sec_num": "3.5"
            },
            {
                "text": "The attached (or dependent) pronouns are used as a suffix added to the verbs ( ex: object pronouns), nouns (ex: possessive pronouns) and prepositions. In terms of the detached (independent) pronouns, they are used as a standalone word and inflected in more than one position in the sentence. Reflexive pronouns in Arabic is similar to that in English, the word \"self\" (\" \", nafs) is used with the attached pronouns to form the reflexive pronouns. Such as ( ) \"nafsahu\" (himself). All pronouns agree to what they refer to in person, gender and number.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Pronouns",
                "sec_num": "3.5"
            },
            {
                "text": "In this section, we describe the Arabic lexicon, list the Arabic language functions proposed and what they cover from the challenges listed in section 3. Table 3 : examples show the personal pronouns (attached and detached).",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 154,
                        "end": 161,
                        "text": "Table 3",
                        "ref_id": null
                    }
                ],
                "eq_spans": [],
                "section": "Design and implementation of the Arabic language functions",
                "sec_num": "4"
            },
            {
                "text": "The lexicon was extracted from Wiktionary 5 . Usually, Wiktionary dump files are parsed to extract information (such as in (Braun et al., 2019) . However, the Arabic dump file is not well formatted. Therefore, a customised web scrap script generated to get all available nouns, verbs and adjectives from the English Wiktionary page. Overall, our lexicon covers around 9000 lemmata (~4000 nouns,~3000 verbs and~2000 adjectives).",
                "cite_spans": [
                    {
                        "start": 123,
                        "end": 143,
                        "text": "(Braun et al., 2019)",
                        "ref_id": "BIBREF8"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Arabic Lexicon",
                "sec_num": "4.1"
            },
            {
                "text": "To cover most of the challenges listed before, six main functions were designed (inflectNoun, buildNP, inflectVerb, inflectAdjective, pronouns and numToWords). Clauses and Pronouns types other than personal pronouns were not covered. The collected lexicon and collections of rules were used to implement these functions based on the information in section 3.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Arabic language functions",
                "sec_num": "4.2"
            },
            {
                "text": "inflectNoun function: It gives the inflected form of the noun in the correct grammatical form. Taking into considerations all aspects number, gender and case. This is done by passing these values based on the sentence, then applying the rules to the targeted noun. The lexicon provides over 4000 nouns inflections, however, if the noun is not found it will be then inflected using an implemented collection of rules. Moreover, the dual and plural forms can be explicitly specified in the function call to get the desired result. (For all functions' definitions and parameters see the Appendices) inflectAdjective function: Same as the noun it can return the right inflected form of the adjective based on the number, gender and case provided using the lexicon and rule-based method. Furthermore, it covers the case of a non-human noun and provides both full and deflected agreement.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Arabic language functions",
                "sec_num": "4.2"
            },
            {
                "text": "numToWords function: It generates the respective words for a given number in two types \"Cardinal\" and \"Ordinal\". Such as 12 will spelt out as ( ) twelve, if the \"Cardinal\" type is used. The generated text follows all the rules of the Arabic numerals mentioned in subsection 3.2.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Arabic language functions",
                "sec_num": "4.2"
            },
            {
                "text": "buildNP (build Noun Phrase) function: It simply generated from the combining of the previous three functions. It will generate a noun phrase with numerals and a list of adjectives inflected correctly. It covers case, gender and number for nouns, numerals and adjectives in addition to the relationship and agreements between them.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Arabic language functions",
                "sec_num": "4.2"
            },
            {
                "text": "inflectVerb function: It returns the inflected form of the verb after applying all the factors that can affect the verb form (number, gender, person, voice, mood and tense). The inflection first searched in the lexicon, if it is not provided then the verb will be inflected using a set of conditions, assuming that the verb belongs to the first form rules. This assumption was made because it is one of the most common forms (Shaalan et al., 2015) and it takes a lot of time and effort to implements all forms for sound and weak verbs.",
                "cite_spans": [
                    {
                        "start": 425,
                        "end": 447,
                        "text": "(Shaalan et al., 2015)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Arabic language functions",
                "sec_num": "4.2"
            },
            {
                "text": "Pronouns function: This covers the personal pronouns. It can produce a correctly inflected pronouns for both attached and detached and for four types (subject, object, reflexive and possessive).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Arabic language functions",
                "sec_num": "4.2"
            },
            {
                "text": "These functions are the first Arabic NLG functions and there are no other similar Arabic approaches to compare with. We evaluated our approach using the same two methods used in (Braun et al., 2019) . The robustness and correctness of the implementation of the Arabic language functions were measured by examining 250 test units consist of different parts of speech (nouns, adjectives, verbs, numerals and pronouns). The results from the functions were all correct, whether it was inflected using a collections of rules or by using the lexical resource.",
                "cite_spans": [
                    {
                        "start": 178,
                        "end": 198,
                        "text": "(Braun et al., 2019)",
                        "ref_id": "BIBREF8"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Evaluation",
                "sec_num": "5"
            },
            {
                "text": "To measure how much these functions can cover from the Arabic language, NADA corpus is used (Alalyani and Marie-Sainte, 2018). The corpus categorised the data into ten categories. An average of 10 sentences extracted from each of the 10 categories provided in NADA and tried to generate them with our language functions. This was successful in 85% of cases.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Evaluation",
                "sec_num": "5"
            },
            {
                "text": "Most of the sentences that the system failed to reproduce were because of the use of other pronouns types or wrong verb inflections when a verb is used in the sentence that is not a form (I) type and not in the lexicon. Such as this sentence \" \" (Karejo Silva scored the game's only goal in the second minute), the verb \" \" (scored) is a singular, past, not form (I) and does not exist in Wiktionary. Another failed sentence is \" \" (The match ended with the victory of Sunderland, who halted Liverpool's balance of 15 points), the word \" \" (who) is a relative pronoun which can take different inflections. Therefore, this sentence considered as failed to reproduce, because relative pronouns are not yet covered by our functions.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Evaluation",
                "sec_num": "5"
            },
            {
                "text": "As mentioned before, as far as we know this is the first NLG system for Arabic. Therefore, we could not compare it with others. However, for a similar approach with different system architecture and German language, Braun et al. 2019achieved 84%, which means our results are acceptable. In terms of scalability, these functions can be adapted easily to use in Classical Arabic and all Arabic dialects (more than 20 dialects) Abdul-Mageed et al. (2018).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Evaluation",
                "sec_num": "5"
            },
            {
                "text": "We plan to enhance the performance of the system by covering other verbs forms, clauses and other types of pronouns (demonstrative and relative pronouns). Furthermore, we intend to increase the number of lemmata covered by the lexicon to get better accuracy, especially with weak verbs and broken plural cases. After that, we can create an Arabic version of simplenlg.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Future Work",
                "sec_num": "6"
            },
            {
                "text": "This work was carried out during the studying duration of the first author for an MSc of AI degree at the University of Aberdeen, which was fully funded by Chevening Scholarships, the UK government's global scholarship programme, funded by the Foreign and Commonwealth Office (FCO) and partner organisations. We also thank Jette Viethen and other people at Arria NLG for their suggestions and help. (default \"1\" for male). It accepts these values for male (1,m,male) and (0,f,female) for female.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Acknowledgments",
                "sec_num": "7"
            },
            {
                "text": "Number (optional) (default \"3\" for the third person). The grammatical person of the verb. Either 1, 2 or 3 to indicate first, second and third person respectively. Voice",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Person",
                "sec_num": null
            },
            {
                "text": "String (optional) (default \"active\"). It accepts only two values \"active\" and \"passive\" for the verb voice.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Person",
                "sec_num": null
            },
            {
                "text": "String (optional) (default \"past\"). It accepts five values (\"past\", \"none_past\", \"subjunctive\", \"jussive\" or \"imperative\"). However, if the value \"passive\" provided in the \"Voice\" parameter or 2nd person is selected then the value \"imperative\" is not acceptable. The input noun. Ideally, the noun in its singular nominative form should be provided, however, if the word did not match any singular nominative form in the DB, then, all other forms will be checked. If the word did not match again, this time the noun will be inflected manually using the known rules of the noun inflections number Number or String (Optional) (default \"plural\" )The grammatical number of the verb. Either singular, dual, plural or a number (not: dual or number 2 cannot be used with the 1st person) Gender Number (optional) (default \"singular\") The grammatical number of the noun. Either singular, dual, plural or a number. Gender Number or string (optional) (default \"1\" for a male if the value, not in the DB). It accepts these values for male (1,m,male) and (0,f,female) for female Case String (optional)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Mood_tense",
                "sec_num": null
            },
            {
                "text": "The default value (nominative). This parameter accepts only these three values \"nominative\", \"accusative\" or \"genitive\". It represents the case of the noun to be inflected base on it.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Mood_tense",
                "sec_num": null
            },
            {
                "text": "String (optional)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Dual",
                "sec_num": null
            },
            {
                "text": "If this value provided, then it overrides the existing \"dual\" value of the noun saved in the database. If not and there is no value in the database for the given noun then it will be treated according to the dual rules explained earlier.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Dual",
                "sec_num": null
            },
            {
                "text": "String (optional)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Plural",
                "sec_num": null
            },
            {
                "text": "If this value provided, then it overrides the existing \"Plural\" value of the noun saved in the database. If not and there is no value in the database for the given noun then it will be treated according to the Sound plural rules explained earlier. A list of adjectives that can be used to describe the noun given in the noun parameter. The full agreement between adjectives and the noun will be used, unless, the value of the \"is_person\" parameter is set to True and the value \"DA\" or False is given in the \"agreement\" parameter.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Plural",
                "sec_num": null
            },
            {
                "text": "Number or String (Optional) (default \"singular\") The grammatical number of the noun. Either singular, dual, plural or a number.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Number",
                "sec_num": null
            },
            {
                "text": "Number or string (optional) (default \"1\" for male). It accepts these values for male (1,m,male) and (0,f,female) for female. This value will be used in case of the adjective is not found in the database only.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Gender",
                "sec_num": null
            },
            {
                "text": "String (optional)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Case",
                "sec_num": null
            },
            {
                "text": "The default value (nominative). This parameter accepts only these three values \"nominative\", \"accusative\" or \"genitive\". It represents the case of the adjective to be inflected base on it Is_person Boolean (optional) (default \"True\"). It indicates whether the modifier(s) refers to a person (human) noun or not. Agreement Boolean (optional) (default \"1\", meaning Full agreement). This parameter now accepts these values (1, \"fa\", \"full\", \"full agreement\") for full agreement with the noun, and (0, \"da\", \"deflected\", \"deflected agreement\") for deflected agreement with the noun in case if the noun has the value \"un-human\" in the database. Table 6 : ar_inflectAdjective parameters",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 640,
                        "end": 647,
                        "text": "Table 6",
                        "ref_id": null
                    }
                ],
                "eq_spans": [],
                "section": "Case",
                "sec_num": null
            },
            {
                "text": "Change the numbers to their respective words in two types \"Cardinal\" and \"Ordinal\". For example: 12 => ( ) twelve, if the \"Cardinal\" type is used. The number to be spelt out.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "ar_numToWords",
                "sec_num": "1.4"
            },
            {
                "text": "The default value (nominative). This parameter accepts only these three values \"nominative\", \"accusative\" or \"genitive\". It represents the case that the number should be inflected base on it.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Case String (optional)",
                "sec_num": null
            },
            {
                "text": "Gender Number or string (optional) (default \"1\" for male). It accepts these values for male (1,m,male) and (0,f,female) for female.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Case String (optional)",
                "sec_num": null
            },
            {
                "text": "String (optional)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Type",
                "sec_num": null
            },
            {
                "text": "Default (cardinal), it can take only two values: \"cardinal\" or \"ordinal\".",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Type",
                "sec_num": null
            },
            {
                "text": "Number or String (optional) (default \"wordsOnly\"). If count is less than or equal to this value it will be spelt out; otherwise, it will be formatted as digits. For example, if you enter the integer 5, then numbers up to and including 5 will be words, and numbers greater than 5 will be digits. \"wordsOnly\", all numbers should be spelt out. \"digitsOnly\", no numbers should be spelt out.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Number Format",
                "sec_num": null
            },
            {
                "text": "String (optional) (default \" \"). The word used to express a value of 0. Either \" \", \" \", or another word or phrase of your choice, e.g. \" \". (default \"singular\") The grammatical number of the pronoun. Either singular, dual, plural or a number (not: dual or number 2 cannot be used with the 1st person, therefore, in this case, the pronoun of the plural first person will be used) Gender Number or string (optional) (default \"1\" for male). It accepts accepts these values for male (1,m,male) and (0,f,female) for female. If this value provided, then it overrides the existing \"Plural\" value of the noun saved in the database. If not and there is no value in the database for the given noun then it will be treated according to the Sound plural rules explained earlier.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Zero Format",
                "sec_num": null
            },
            {
                "text": "Is person Boolean (optional) (default \"True\") indicate if the word provided in the \"word\" parameter refers to a human or not. Because, for nonhuman plurals in the Arabic language, the singular feminine pronoun in its third-person form is used (Mandatory) Either \"dual\", \"plural\" or a number. If a number is provided, it will be output as a modifier to the input noun and the noun is correctly inflected. If singular, dual or plural is used, the noun is inflected correctly based on the noun gender and the \"case\" parameter value.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Zero Format",
                "sec_num": null
            },
            {
                "text": "The input noun. Ideally, the noun in its singular nominative form should be provided, however, if the word did not match any singular nominative form in the DB, then, all other forms will be checked. If the word did not match again, this time the noun will be inflected manually using the known rules of the noun inflections.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "String (Mandatory)",
                "sec_num": null
            },
            {
                "text": "String (optional)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Case",
                "sec_num": null
            },
            {
                "text": "The default value (nominative). This parameter accepts only these three values \"nominative\", \"accusative\" or \"genitive\". It represents the case of the noun to be inflected base on it. Gender Number or string (optional) (default \"1\" for male). It accepts these values for male (1,m,male) and (0,f,female) for female. If this value provided, then it overrides the existing gender of the noun saved in the database. If not and there is no value in the database for the given noun then it will be treated as \"male\" noun.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Case",
                "sec_num": null
            },
            {
                "text": "String (optional)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Dual",
                "sec_num": null
            },
            {
                "text": "If this value provided, then it overrides the existing \"dual\" value of the noun saved in the database. If not and there is no value in the database for the given noun then it will be treated according to the dual rules explained earlier.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Dual",
                "sec_num": null
            },
            {
                "text": "String (optional)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Plural",
                "sec_num": null
            },
            {
                "text": "If this value provided, then it overrides the existing \"Plural\" value of the noun saved in the database. If not and there is no value in the database for the given noun then it will be treated according to the Sound plural rules explained earlier.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Plural",
                "sec_num": null
            },
            {
                "text": "Number or String (optional) (default \"wordsOnly\"). If count is less than or equal to this value it will be spelt out; otherwise, it will be formatted as digits. For example, if you enter the integer 5, then numbers up to and including 5 will be words, and numbers greater than 5 will be digits. \"wordsOnly\", all numbers should be spelt out. \"digitsOnly\", no numbers should be spelt out Zero Format",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Number Format",
                "sec_num": null
            },
            {
                "text": "String (optional) (default \" \"). The word used to express a value of 0. Either \" \", \" \", or another word or phrase of your choice, e.g. \"",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Number Format",
                "sec_num": null
            },
            {
                "text": "\".",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Number Format",
                "sec_num": null
            },
            {
                "text": "A list of adjectives that can be used to describe the noun given in the noun parameter. The full agreement between adjectives and the noun will be used, unless, the noun available in the database and has the value \"un-human\" and the value \"DA\" is given in the \"agreement\" parameter. Agreement Boolean (optional) (default \"1\", meaning Full agreement). This parameter now accepts these values (1, \"fa\", \"full\", \"full agreement\") for full agreement with the noun, and (0, \"da\", \"deflected\", \"deflected agreement\") for deflected agreement with the noun in case if the noun has the value \"un-human\" in the database. ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Modifiers List of Strings (Optional)",
                "sec_num": null
            },
            {
                "text": "https://github.com/WaelMohammedAbed/ Natural-Language-Generation-for-the-Arabic-Language",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "https://rosaenlg.org/rosaenlg/1.16.4/ about/compare.html 3 https://github.com/societe-generale/ core-nlg 4 https://xspin.it/plus",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "https://en.wiktionary.org/wiki/ Category:Arabic_lemmas",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            }
        ],
        "back_matter": [],
        "bib_entries": {
            "BIBREF0": {
                "ref_id": "b0",
                "title": "Lessons 31-45; Appendices",
                "authors": [
                    {
                        "first": "W",
                        "middle": [
                            "M"
                        ],
                        "last": "Zn Abdel-Malek",
                        "suffix": ""
                    },
                    {
                        "first": "G",
                        "middle": [
                            "N"
                        ],
                        "last": "Erwin",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Saad",
                        "suffix": ""
                    }
                ],
                "year": 1983,
                "venue": "",
                "volume": "2",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "ZN Abdel-Malek, WM Erwin, and GN Saad. 1983. Elementary Modern Standard Arabic: Volume 2, Lessons 31-45; Appendices, volume 2. Cambridge University Press.",
                "links": null
            },
            "BIBREF1": {
                "ref_id": "b1",
                "title": "You tweet what you speak: A city-level dataset of arabic dialects",
                "authors": [
                    {
                        "first": "Muhammad",
                        "middle": [],
                        "last": "Abdul-Mageed",
                        "suffix": ""
                    },
                    {
                        "first": "Hassan",
                        "middle": [],
                        "last": "Alhuzali",
                        "suffix": ""
                    },
                    {
                        "first": "Mohamed",
                        "middle": [],
                        "last": "Elaraby",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proceedings of the Eleventh International Conference on Language Resources and Evaluation",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Muhammad Abdul-Mageed, Hassan Alhuzali, and Mo- hamed Elaraby. 2018. You tweet what you speak: A city-level dataset of arabic dialects. In Proceedings of the Eleventh International Conference on Lan- guage Resources and Evaluation (LREC 2018).",
                "links": null
            },
            "BIBREF2": {
                "ref_id": "b2",
                "title": "The role of short vowels in reading arabic: A critical literature review",
                "authors": [
                    {
                        "first": "Salim",
                        "middle": [],
                        "last": "Abu-Rabia",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Journal of psycholinguistic research",
                "volume": "48",
                "issue": "4",
                "pages": "785--795",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Salim Abu-Rabia. 2019. The role of short vowels in reading arabic: A critical literature review. Journal of psycholinguistic research, 48(4):785-795.",
                "links": null
            },
            "BIBREF3": {
                "ref_id": "b3",
                "title": "The syntax of (complex)",
                "authors": [
                    {
                        "first": "Hussein",
                        "middle": [],
                        "last": "Al",
                        "suffix": ""
                    },
                    {
                        "first": "-",
                        "middle": [],
                        "last": "Bataineh",
                        "suffix": ""
                    },
                    {
                        "first": "Phil",
                        "middle": [],
                        "last": "Branigan",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Hussein Al-Bataineh and Phil Brani- gan. 2020. The syntax of (complex)",
                "links": null
            },
            "BIBREF4": {
                "ref_id": "b4",
                "title": "Noun quantification in english and modern standard arabic: a descriptive morphological comparison",
                "authors": [],
                "year": 1997,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "DOI": [
                        "10.31219/osf.io/6pesa"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "https://doi.org/10.31219/osf.io/6pesa. Noor Sultan Al Easa. 1997. Noun quantifica- tion in english and modern standard ara- bic: a descriptive morphological comparison. Http://hdl.handle.net/10576/8069.",
                "links": null
            },
            "BIBREF5": {
                "ref_id": "b5",
                "title": "Applications of mining arabic text: A review",
                "authors": [
                    {
                        "first": "Qasem",
                        "middle": [],
                        "last": "Al-Radaideh",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Recent Trends in Computational Intelligence. IntechOpen",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Qasem Al-Radaideh. 2020. Applications of mining ara- bic text: A review. In Recent Trends in Computa- tional Intelligence. IntechOpen.",
                "links": null
            },
            "BIBREF6": {
                "ref_id": "b6",
                "title": "Nada: New arabic dataset for text classification",
                "authors": [
                    {
                        "first": "Nada",
                        "middle": [],
                        "last": "Alalyani",
                        "suffix": ""
                    },
                    {
                        "first": "Souad Larabi Marie-Sainte",
                        "middle": [],
                        "last": "",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "International Journal of Advanced Computer Science and Applications",
                "volume": "9",
                "issue": "9",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Nada Alalyani and Souad Larabi Marie-Sainte. 2018. Nada: New arabic dataset for text classification. In- ternational Journal of Advanced Computer Science and Applications, 9(9).",
                "links": null
            },
            "BIBREF7": {
                "ref_id": "b7",
                "title": "Google ngram viewer does not include arabic corpus! towards n-gram viewer for arabic corpus",
                "authors": [
                    {
                        "first": "Izzat",
                        "middle": [],
                        "last": "Alsmadi",
                        "suffix": ""
                    },
                    {
                        "first": "Mohammad",
                        "middle": [],
                        "last": "Zarour",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Int. Arab J. Inf. Technol",
                "volume": "15",
                "issue": "5",
                "pages": "785--794",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Izzat Alsmadi and Mohammad Zarour. 2018. Google n- gram viewer does not include arabic corpus! towards n-gram viewer for arabic corpus. Int. Arab J. Inf. Technol., 15(5):785-794.",
                "links": null
            },
            "BIBREF8": {
                "ref_id": "b8",
                "title": "Simplenlg-de: Adapting simplenlg 4 to german",
                "authors": [
                    {
                        "first": "Daniel",
                        "middle": [],
                        "last": "Braun",
                        "suffix": ""
                    },
                    {
                        "first": "Kira",
                        "middle": [],
                        "last": "Klimt",
                        "suffix": ""
                    },
                    {
                        "first": "Daniela",
                        "middle": [],
                        "last": "Schneider",
                        "suffix": ""
                    },
                    {
                        "first": "Florian",
                        "middle": [],
                        "last": "Matthes",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 12th International Conference on Natural Language Generation",
                "volume": "",
                "issue": "",
                "pages": "415--420",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Daniel Braun, Kira Klimt, Daniela Schneider, and Flo- rian Matthes. 2019. Simplenlg-de: Adapting sim- plenlg 4 to german. In Proceedings of the 12th Inter- national Conference on Natural Language Genera- tion, pages 415-420.",
                "links": null
            },
            "BIBREF9": {
                "ref_id": "b9",
                "title": "Adapting simplenlg to galician language",
                "authors": [
                    {
                        "first": "Andrea",
                        "middle": [],
                        "last": "Cascallar-Fuentes",
                        "suffix": ""
                    },
                    {
                        "first": "Alejandro",
                        "middle": [],
                        "last": "Ramos-Soto",
                        "suffix": ""
                    },
                    {
                        "first": "Alberto",
                        "middle": [
                            "Bugar\u00edn"
                        ],
                        "last": "Diz",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proceedings of the 11th International Conference on Natural Language Generation",
                "volume": "",
                "issue": "",
                "pages": "67--72",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Andrea Cascallar-Fuentes, Alejandro Ramos-Soto, and Alberto Bugar\u00edn Diz. 2018. Adapting simplenlg to galician language. In Proceedings of the 11th Inter- national Conference on Natural Language Genera- tion, pages 67-72.",
                "links": null
            },
            "BIBREF10": {
                "ref_id": "b10",
                "title": "Simplenlg-zh: a linguistic realisation engine for mandarin",
                "authors": [
                    {
                        "first": "Guanyi",
                        "middle": [],
                        "last": "Chen",
                        "suffix": ""
                    },
                    {
                        "first": "Kees",
                        "middle": [],
                        "last": "Van Deemter",
                        "suffix": ""
                    },
                    {
                        "first": "Chenghua",
                        "middle": [],
                        "last": "Lin",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proceedings of the 11th International Conference on Natural Language Generation",
                "volume": "",
                "issue": "",
                "pages": "57--66",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Guanyi Chen, Kees Van Deemter, and Chenghua Lin. 2018. Simplenlg-zh: a linguistic realisation engine for mandarin. In Proceedings of the 11th Interna- tional Conference on Natural Language Generation, pages 57-66.",
                "links": null
            },
            "BIBREF11": {
                "ref_id": "b11",
                "title": "Learnability and generalisation of arabic broken plural nouns. Language, cognition and neuroscience",
                "authors": [
                    {
                        "first": "Lisa",
                        "middle": [],
                        "last": "Garnand Dawdy-Hesterberg",
                        "suffix": ""
                    },
                    {
                        "first": "Janet",
                        "middle": [
                            "Breckenridge"
                        ],
                        "last": "Pierrehumbert",
                        "suffix": ""
                    }
                ],
                "year": 2014,
                "venue": "",
                "volume": "29",
                "issue": "",
                "pages": "1268--1282",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Lisa Garnand Dawdy-Hesterberg and Janet Brecken- ridge Pierrehumbert. 2014. Learnability and gener- alisation of arabic broken plural nouns. Language, cognition and neuroscience, 29(10):1268-1282.",
                "links": null
            },
            "BIBREF12": {
                "ref_id": "b12",
                "title": "A statistical, grammar-based approach to microplanning",
                "authors": [
                    {
                        "first": "Claire",
                        "middle": [],
                        "last": "Gardent",
                        "suffix": ""
                    },
                    {
                        "first": "Laura",
                        "middle": [],
                        "last": "Perez-Beltrachini",
                        "suffix": ""
                    }
                ],
                "year": 2017,
                "venue": "Computational Linguistics",
                "volume": "43",
                "issue": "1",
                "pages": "1--30",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Claire Gardent and Laura Perez-Beltrachini. 2017. A statistical, grammar-based approach to microplan- ning. Computational Linguistics, 43(1):1-30.",
                "links": null
            }
        },
        "ref_entries": {
            "TABREF1": {
                "type_str": "table",
                "content": "<table><tr><td>: examples show the Arabic noun's forms</td></tr><tr><td>with different case and gender</td></tr><tr><td>on the case and the gender of the noun.</td></tr></table>",
                "html": null,
                "text": "",
                "num": null
            },
            "TABREF3": {
                "type_str": "table",
                "content": "<table><tr><td>: examples show the numerals inflection</td></tr><tr><td>1997).</td></tr></table>",
                "html": null,
                "text": "",
                "num": null
            },
            "TABREF5": {
                "type_str": "table",
                "content": "<table><tr><td/><td colspan=\"3\">A Appendix 1 -Definitions and</td></tr><tr><td/><td colspan=\"3\">parameters of the Arabic language</td></tr><tr><td/><td colspan=\"2\">functions</td><td/></tr><tr><td>Albert Gatt and Ehud Reiter. 2009. Simplenlg: A re-</td><td colspan=\"3\">1.1 ar_inflectVerb</td></tr><tr><td>alisation engine for practical applications. In Pro-ceedings of the 12th European Workshop on Natural Language Generation (ENLG 2009), pages 90-93.</td><td colspan=\"3\">Returns the inflected form of the input verb taking into consideration five aspects (number, gender, person, voice and mood/tense).</td></tr><tr><td>Spence Green and John DeNero. 2012. A class-based</td><td/><td/><td/></tr><tr><td>agreement model for generating accurately inflected translations. In Proceedings of the 50th Annual</td><td>Parameter Word</td><td>Type string</td><td>description (Mandatory) The input verb. Ideally, the verb root should be provided, however, if the word did not</td></tr><tr><td>Meeting of the Association for Computational Lin-guistics (Volume 1: Long Papers), pages 146-155.</td><td/><td/><td>match any root in the DB, then, all other forms will be checked. If the word did not match again, this time the verb will be inflected manually using the</td></tr><tr><td/><td/><td/><td>known rules of the verb inflections</td></tr><tr><td>Ali Idrissi, Eiman Mustafawi, Tariq Khwaileh, and R Muralikrishnan. 2019. A neurophysiological</td><td>number</td><td>Number or String (Op-</td><td>(default \"plural\" )The grammatical number of the verb. Either singular, dual, plural or a number (not:</td></tr><tr><td>study of noun-adjective agreement in arabic: The</td><td>Gender</td><td>tional) Number</td><td>dual or number 2 cannot be used with the 1st person)</td></tr><tr><td>impact of animacy and diglossia on the dynamics of language processing. bioRxiv, page 729855.</td><td/><td>or (optional) string</td><td/></tr><tr><td>Zainab Kadim Igaab and Intisar Raham Tarrad. 2019.</td><td/><td/><td/></tr><tr><td>Pronouns in english and arabic: a contrastive study.</td><td/><td/><td/></tr><tr><td>Engl. Lang. Lit. Stud, 9(1).</td><td/><td/><td/></tr><tr><td>Ruud de Jong and Mari\u00ebt Theune. 2018. Going dutch:</td><td/><td/><td/></tr><tr><td>Creating simplenlg-nl. In Proceedings of the 11th</td><td/><td/><td/></tr><tr><td>International Conference on Natural Language Gen-</td><td/><td/><td/></tr><tr><td>eration, pages 73-78.</td><td/><td/><td/></tr><tr><td>Nabil Khoufi, Chafik Aloulou, and Lamia Hadrich Bel-</td><td/><td/><td/></tr><tr><td>guith. 2016. Parsing arabic using induced probabilis-</td><td/><td/><td/></tr><tr><td>tic context free grammar. International Journal of</td><td/><td/><td/></tr><tr><td>Speech Technology, 19(2):313-323.</td><td/><td/><td/></tr><tr><td>Alessandro Mazzei, Cristina Battaglino, and Cristina</td><td/><td/><td/></tr><tr><td>Bosco. 2016. Simplenlg-it: adapting simplenlg to</td><td/><td/><td/></tr><tr><td>italian. In Proceedings of the 9th international natu-</td><td/><td/><td/></tr><tr><td>ral language generation conference, pages 184-192.</td><td/><td/><td/></tr><tr><td>Yael Netzer and Michael Elhadad. 1998. Generation of</td><td/><td/><td/></tr><tr><td>noun compounds in hebrew: Can syntactic knowl-</td><td/><td/><td/></tr><tr><td>edge be fully encapsulated? In Natural Language</td><td/><td/><td/></tr><tr><td>Generation.</td><td/><td/><td/></tr><tr><td>Alejandro Ramos-Soto, Julio Janeiro-Gallardo, and Al-</td><td/><td/><td/></tr><tr><td>berto Bugar\u00edn Diz. 2017. Adapting simplenlg to</td><td/><td/><td/></tr><tr><td>spanish. In Proceedings of the 10th international</td><td/><td/><td/></tr><tr><td>conference on natural language generation, pages</td><td/><td/><td/></tr><tr><td>144-148.</td><td/><td/><td/></tr><tr><td>Karin C Ryding. 2005. A reference grammar of mod-</td><td/><td/><td/></tr><tr><td>ern standard Arabic. Cambridge university press.</td><td/><td/><td/></tr><tr><td>Khaled Shaalan, Marwa Magdy, and Aly Fahmy. 2015.</td><td/><td/><td/></tr><tr><td>Analysis and feedback of erroneous arabic verbs.</td><td/><td/><td/></tr><tr><td>Natural Language Engineering, 21(2):271.</td><td/><td/><td/></tr><tr><td>Muayad Abdul-Halim Ahmad Shamsan and Abdul-</td><td/><td/><td/></tr><tr><td>majeed Attayib. 2015. Inflectional morphology in</td><td/><td/><td/></tr><tr><td>arabic and english: a contrastive study. Interna-</td><td/><td/><td/></tr><tr><td>tional Journal of English Linguistics, 5(2):139.</td><td/><td/><td/></tr></table>",
                "html": null,
                "text": "Albert Gatt and Emiel Krahmer. 2018. Survey of the state of the art in natural language generation: Core tasks, applications and evaluation. Journal of Artificial Intelligence Research, 61:65-170.",
                "num": null
            },
            "TABREF6": {
                "type_str": "table",
                "content": "<table><tr><td/><td colspan=\"2\">: ar_inflectVerb parameters</td></tr><tr><td colspan=\"3\">1.2 ar_inflectNoun</td></tr><tr><td colspan=\"3\">Returns the inflected form of the input noun taking</td></tr><tr><td colspan=\"3\">into consideration 3 aspects (number, gender and</td></tr><tr><td>case).</td><td/><td/></tr><tr><td>Parameter</td><td>Type</td><td>description</td></tr><tr><td>Word</td><td>string</td><td/></tr><tr><td/><td>(Manda-</td><td/></tr><tr><td/><td>tory)</td><td/></tr></table>",
                "html": null,
                "text": "",
                "num": null
            },
            "TABREF7": {
                "type_str": "table",
                "content": "<table><tr><td>: ar_inflectNoun parameters</td></tr><tr><td>1.3 ar_inflectAdjective</td></tr><tr><td>Returns the inflected form of the input adjective</td></tr><tr><td>taking into consideration 3 aspects (number, gender</td></tr><tr><td>and case).</td></tr></table>",
                "html": null,
                "text": "",
                "num": null
            },
            "TABREF9": {
                "type_str": "table",
                "content": "<table><tr><td>: ar_numToWords parameters</td></tr><tr><td>1.5 ar_pronouns</td></tr><tr><td>Returns the pronoun (e.g. you, his, her, them, etc.)</td></tr><tr><td>for a given input gender or name. The optional pa-</td></tr><tr><td>rameters allow you to specify your desired pronoun</td></tr><tr><td>type.</td></tr></table>",
                "html": null,
                "text": "",
                "num": null
            },
            "TABREF11": {
                "type_str": "table",
                "content": "<table><tr><td>: ar_pronouns parameters</td></tr></table>",
                "html": null,
                "text": "",
                "num": null
            },
            "TABREF12": {
                "type_str": "table",
                "content": "<table/>",
                "html": null,
                "text": "",
                "num": null
            }
        }
    }
}