| { | |
| "paper_id": "2020", | |
| "header": { | |
| "generated_with": "S2ORC 1.0.0", | |
| "date_generated": "2023-01-19T07:28:14.435874Z" | |
| }, | |
| "title": "SimpleNLG-TI: Adapting SimpleNLG to Tibetan", | |
| "authors": [ | |
| { | |
| "first": "Zewang", | |
| "middle": [], | |
| "last": "Kuanzhuo", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Tibetan information processing and Machine Translation Key Laboratory of Qinghai Province", | |
| "institution": "Qinghai Normal University", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Lin", | |
| "middle": [], | |
| "last": "Li", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "Qinghai Normal University", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Weina", | |
| "middle": [], | |
| "last": "Zhao", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "Qinghai Normal University", | |
| "location": {} | |
| }, | |
| "email": "" | |
| } | |
| ], | |
| "year": "", | |
| "venue": null, | |
| "identifiers": {}, | |
| "abstract": "Surface realisation is the last but not the least phase of Natural Language Generation, which aims to produce high-quality natural language text based on meaning representations. In this article, we present our work on SimpleNLG-TI, a Tibetan surface realiser, which follows the design paradigm of SimpleNLG-EN. SimpleNLG-TI is built up by our investigation of the core features of Tibetan morphology and syntax. Through this work, we provide a robust and flexible surface realiser for Tibetan generation systems.", | |
| "pdf_parse": { | |
| "paper_id": "2020", | |
| "_pdf_hash": "", | |
| "abstract": [ | |
| { | |
| "text": "Surface realisation is the last but not the least phase of Natural Language Generation, which aims to produce high-quality natural language text based on meaning representations. In this article, we present our work on SimpleNLG-TI, a Tibetan surface realiser, which follows the design paradigm of SimpleNLG-EN. SimpleNLG-TI is built up by our investigation of the core features of Tibetan morphology and syntax. Through this work, we provide a robust and flexible surface realiser for Tibetan generation systems.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Abstract", | |
| "sec_num": null | |
| } | |
| ], | |
| "body_text": [ | |
| { | |
| "text": "Natural Language Generation(NLG) aims to produces Natural Language based on various input like meaning representations or numeric data. NLG systems can be classified into text-to-text or data-to-text system according to the system input. Most early generation work in Tibetan focused on text-to-text generation by Neural Networks, for instance, Poem Generation (Chajia, 2019) , Machine Translation (Yachao, 2017) , Question Answering (Sun et al., 2019; Xiaoying, 2017) . There is still a lack of theoretical studies and available resources for Tibetan generation research. To solve the latter problem, this paper offers a practical tool, SimpleNLG-TI, for generating Tibetan text.", | |
| "cite_spans": [ | |
| { | |
| "start": 361, | |
| "end": 375, | |
| "text": "(Chajia, 2019)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 398, | |
| "end": 412, | |
| "text": "(Yachao, 2017)", | |
| "ref_id": "BIBREF20" | |
| }, | |
| { | |
| "start": 434, | |
| "end": 452, | |
| "text": "(Sun et al., 2019;", | |
| "ref_id": "BIBREF16" | |
| }, | |
| { | |
| "start": 453, | |
| "end": 468, | |
| "text": "Xiaoying, 2017)", | |
| "ref_id": "BIBREF19" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "Classical pipeline model (Reiter and Dale, 1997) proposed that a generation process shall accomplish six tasks, according to which the last but not the least phase of the process is surface realisation. Surface realisation aims to produce strings that conform to grammar rules, therefore it plays an important role on improving readability of output. To build up SimpleNLG-TI, we solve key theoretical and practical problems in Tibetan realising. Firstly, we build up a linguistic rule bank for SimpleNLG-TI by exploring morphology and syntax features in Tibetan such as function words, inflection, word order. Secondly, in order to develop a practical surface realiser we investigate how to apply the framework of SimpleNLG-EN to SimpleNLG-TI properly and reuse its code 1 .", | |
| "cite_spans": [ | |
| { | |
| "start": 25, | |
| "end": 48, | |
| "text": "(Reiter and Dale, 1997)", | |
| "ref_id": "BIBREF13" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "The rest of this paper is organised as follows. In section 2, we discuss some related work. In section 3, we discuss Tibetan morphosyntactic characteristics and how to apply them into SimpleNLG-TI. Evaluation process is introduced in section 4. And we end with conclusion and future work.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "As an important member of Sino-Tibetan language, Tibetan has 3 major dialect areas in China and is used by about 8 million people around the world. A classic work(de K\u00f6r\u00f6s, 1834) analyzed phonetic and morphology, and syntax features of Tibetan. A primary feature of Tibetan is its SOV word-order, for instance, in the predicate of example (1) \u0f5f (eat) locates at the end of the sentence. Moreover, there are abundant reductions in Tibetan. Both the original case \u0f61\u0f72\u0f66 and its reduction form \u0f66 are correct in example (1). In most situation, the reduction form is preferred.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Related Work", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "(1) Tibetan: \u0f44\u0f66\u0f0b\u0f64\u0f0b\u0f5f\u0f0d (= \u0f44\u0f0b\u0f61\u0f72\u0f66\u0f0b\u0f64\u0f0b\u0f5f\u0f0d) English: I eat meat. In this work, the coverage of SimpleNLG-TI mainly refer to two highly-respected grammar books (Jumian, 1987; de K\u00f6r\u00f6s, 1834) , which offer us clear and operable grammar. Besides of the earlier work of linguistics, rich language resources have been collected for studies in data-driven Tibetan processing, for instance, a large scale Tibetan corpus has been built up (Liu et al., 2012) which provides SimpleNLG-TI with an available vocabulary.", | |
| "cite_spans": [ | |
| { | |
| "start": 151, | |
| "end": 165, | |
| "text": "(Jumian, 1987;", | |
| "ref_id": "BIBREF7" | |
| }, | |
| { | |
| "start": 166, | |
| "end": 181, | |
| "text": "de K\u00f6r\u00f6s, 1834)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 423, | |
| "end": 441, | |
| "text": "(Liu et al., 2012)", | |
| "ref_id": "BIBREF11" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Related Work", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "As a significant phase of NLG, surface realisation has attracted much research attention and many surface realisers have been developed such as KPM (Bateman, 1997) , YAG(McRoy et al., 2000) , and OPENCCG (White, 2006) . SimpleNLG-EN is an open source project (Gatt and Reiter, 2009,?) , which allows it to be integrated into various language generation applications. For instance, in sentence generating(Sheikha and Inkpen, 2011) and sentence summarizing (Khan et al., 2015) . To realise well-formed text, SimpleNLG-EN adopts rule-based approach based on lexicon and rules with rich linguistic knowledge. The paradigm of Sim-pleNLG has been successfully implemented into many natural language such as Chinese SimpleNLG-ZH (Chen et al., 2018) and German SimpleNLG-DE (Braun et al., 2019) , and SimpleNLG-EnFr (Vaudry and Lapalme, 2013) .", | |
| "cite_spans": [ | |
| { | |
| "start": 148, | |
| "end": 163, | |
| "text": "(Bateman, 1997)", | |
| "ref_id": "BIBREF0" | |
| }, | |
| { | |
| "start": 166, | |
| "end": 189, | |
| "text": "YAG(McRoy et al., 2000)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 204, | |
| "end": 217, | |
| "text": "(White, 2006)", | |
| "ref_id": "BIBREF18" | |
| }, | |
| { | |
| "start": 259, | |
| "end": 284, | |
| "text": "(Gatt and Reiter, 2009,?)", | |
| "ref_id": "BIBREF5" | |
| }, | |
| { | |
| "start": 455, | |
| "end": 474, | |
| "text": "(Khan et al., 2015)", | |
| "ref_id": "BIBREF8" | |
| }, | |
| { | |
| "start": 722, | |
| "end": 741, | |
| "text": "(Chen et al., 2018)", | |
| "ref_id": "BIBREF3" | |
| }, | |
| { | |
| "start": 766, | |
| "end": 786, | |
| "text": "(Braun et al., 2019)", | |
| "ref_id": "BIBREF1" | |
| }, | |
| { | |
| "start": 808, | |
| "end": 834, | |
| "text": "(Vaudry and Lapalme, 2013)", | |
| "ref_id": "BIBREF17" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Related Work", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "SimpleNLG-TI employs the framework of SimpleNLG-EN because of its comprehensive linguistic component definition. In process of designing and developing SimpleNLG-TI, we firstly build up a component set contains content words, function words, and highlyfrequent phrases. Then, to realise linguistic components into a well-formed string, we also establish a rule bank includes orthography, morphology and syntax rules by making use of linguistic knowledge in Tibetan grammar books.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Coverage of SimpleNLG-TI", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "The rich morphological features of Tibetan lead to challenge for surface realisation task. In this work, we only focus on core morphological and syntactic characteristics that Input Code lexicon = XMLLexicon() nlgFactory = NLGFactory(lexicon) sentence = nlgFactory.createClause() sentence.setSubject(\"\u0f64\u0f72\u0f44\u0f0b\u0f58\u0f41\u0f53\u0f0b\") sentence.addCase(\"\u0f42\u0fb1\u0f72 \u0f66\u0f0b\") sentence.addComplement(\"\u0f63\u0f9f\u0f0b\u0f62\u0f7a\u0f66\u0f0b\u0f61\u0f72\u0f66\u0f0b\") sentence.setVerb(\"\u0f56\u0f45\u0f51\u0f0b\") output = Realiser.realiseSentence(sentence) play important roles on realisation. Table 1 shows how generates the following sentence by SimpleNLG-TI. Tibetan: \u0f64\u0f72\u0f44\u0f0b\u0f58\u0f41\u0f53\u0f0b\u0f42\u0fb1\u0f72 \u0f66\u0f0b\u0f64\u0f72\u0f44\u0f0b\u0f63\u0f9f\u0f0b\u0f62\u0f7a\u0f66\u0f0b\u0f61\u0f72\u0f66\u0f0b\u0f56\u0f45\u0f51\u0f0d English: A carpenter cut down a tree with axe.", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 483, | |
| "end": 490, | |
| "text": "Table 1", | |
| "ref_id": "TABREF0" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Coverage of SimpleNLG-TI", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "Nouns and verbs are morphology varied in Tibetan, therefore SimpleNLG-TI takes inflections of nouns and verbs into account because of its significant role in surface realising.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Morphology", | |
| "sec_num": "3.1" | |
| }, | |
| { | |
| "text": "In Tibetan, the concept of plural is realised by using various plural marks like \u0f62\u0fa3\u0f58\u0f66, \u0f5a\u0f7c, \u0f45\u0f42, \u0f51\u0f42, \u0f58\u0f50\u0f60\u0f0b\u0f51\u0f42, \u0f40\u0f74\u0f53, \u0f50\u0f58\u0f66\u0f0b\u0f45\u0f51, \u0f61\u0f7c\u0f44\u0f66, and \u0f66\u0f7c\u0f42\u0f66. SimpleNLG-TI can realise four commonly used plural marks, i.e., \u0f62\u0fa3\u0f58\u0f66, \u0f45\u0f42, \u0f5a\u0f7c, and \u0f51\u0f42. The first three marks can be used with most nouns, yet the last one usually only following inanimate nouns. We provide several examples of different usages of plural marks as followed.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Collective markers", | |
| "sec_num": "3.1.1" | |
| }, | |
| { | |
| "text": "Example (2) shows a common noun student with tow different plural marks \u0f62\u0fa3\u0f58\u0f66 and \u0f51\u0f42 respectively, both of which express the meaning of students. An example of inanimate noun table is made in example (3). example (4) and (5) are instances of how to expressing plural concept of pronouns by \u0f5a\u0f7c and \u0f45\u0f42.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Collective markers", | |
| "sec_num": "3.1.1" | |
| }, | |
| { | |
| "text": "(2) \uf5d4\u0f7c\u0f56\u0f0b\u0f58(student) \uf5d4\u0f7c\u0f56\u0f0b\u0f58\u0f0b\u0f62\u0fa3\u0f58\u0f66(students) \uf5d4\u0f7c\u0f56\u0f0b\u0f58(student) \uf5d4\u0f7c\u0f56\u0f0b\u0f58\u0f0b\u0f51\u0f42(students) ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Collective markers", | |
| "sec_num": "3.1.1" | |
| }, | |
| { | |
| "text": "A simple sentence is normally composed of nouns (or noun phrases), verbs (or verb phrases), and function words such as case and aspect. In this section, we discuss word order in Tibetan briefly, then explain how SimpleNLG-TI realises noun and verb phrases.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Syntax", | |
| "sec_num": "3.2" | |
| }, | |
| { | |
| "text": "Primary word order in Tibetan is Subject-Object-Verb(SOV), which means predicate normally locates at the end of a sentence. A predicate is normally composed of a verb (or verb phrase) or an adjective (or adjective phrase) (Li and Long, 2012) . Example (8) is a simple sentence which is composed of a noun and verb phrase. Take the process of realising (8) as an example, SimpleNLG-TI firstly realises the noun as an independent component, and then separately generates the elements of the verb phrase as an object and complement. The sentence mentioned earlier in this section can be realised by SimpleNLG-TI as sentence (9). In addition to realising a noun phrase and a verb phrase in the same way of (8), the case mark \u0f42\u0fb1\u0f72 \u0f66 is an essential input.", | |
| "cite_spans": [ | |
| { | |
| "start": 222, | |
| "end": 241, | |
| "text": "(Li and Long, 2012)", | |
| "ref_id": "BIBREF10" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Word order", | |
| "sec_num": "3.2.1" | |
| }, | |
| { | |
| "text": "(8) Tibetan: \u0f44\u0f0b\u0f51\u0f42\u0f7a\u0f0b\u0f62\u0f92\u0f53\u0f0b\u0f0b\u0f61\u0f72\u0f53\u0f0d English: I teacher am (9) Tibetan: \u0f64\u0f72\u0f44\u0f0b\u0f58\u0f41\u0f53\u0f0b\u0f42\u0fb1\u0f72 \u0f66\u0f0b\u0f64\u0f72\u0f44\u0f0b\u0f63\u0f9f\u0f0b\u0f62\u0f7a\u0f66\u0f0b\u0f61\u0f72\u0f66\u0f0b\u0f56\u0f45\u0f51\u0f0d English: The carpenter the tree with an axe cut down.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Word order", | |
| "sec_num": "3.2.1" | |
| }, | |
| { | |
| "text": "Negation words are \u0f58, \u0f58\u0f72, \u0f58\u0f7a\u0f51, and \u0f58\u0f72\u0f53, which modify nouns, verbs and adjectives. In this work, we follow investigations proposed by (Jieben, 2012) , thus the rules of negation words using by SimpleNLG-TI are listed here.", | |
| "cite_spans": [ | |
| { | |
| "start": 133, | |
| "end": 147, | |
| "text": "(Jieben, 2012)", | |
| "ref_id": "BIBREF6" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Negation", | |
| "sec_num": "3.2.2" | |
| }, | |
| { | |
| "text": "1. Both \u0f58 and \u0f58\u0f72 are treated as prepositional negatives, which are used to modify verbs and adjectives.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Negation", | |
| "sec_num": "3.2.2" | |
| }, | |
| { | |
| "text": "2. Besides of POS (i.e., part-of-speech), tense also determines the choice of negation words. In SimpleNLG-TI, \u0f58\u0f72 is usually used in present and future tense like (10) and \u0f58 often used in past tense such as example (11). ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Negation", | |
| "sec_num": "3.2.2" | |
| }, | |
| { | |
| "text": "SimpleNLG-TI is capable of realising a noun phrase whose head word is modified by an adjective or noun. When modification is an adjective, SimpleNLG-TI can realised a noun phrase in two ways:", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Noun phrase and Adjective phrase", | |
| "sec_num": "3.2.3" | |
| }, | |
| { | |
| "text": "1. noun + noun/adjective like example (13) 2. adjective/pronoun + case + noun such as example 14The realisation of an adjective phrase follows the way of noun phrases, for instance, example (15). The adjective phrase is realized as a noun phrase, the adjective is the center word which is modified by the adverb. ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Noun phrase and Adjective phrase", | |
| "sec_num": "3.2.3" | |
| }, | |
| { | |
| "text": "Structure of verb phrases is varied, which is partial because the head words can be simultaneously modified by various components such as auxiliary, adverb, etc. According to whether a verb phrase contains case or not, SimpleNLG-TI realises verb phrases as follows. 1. noun + verb: in example 16and 172. noun + (case) + verb: object-verb structure realising such as example 18 ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Verb phrase", | |
| "sec_num": "3.2.4" | |
| }, | |
| { | |
| "text": "SimpleNLG-TI can realise two punctuation, that is vertical symbol \u0f0d and phonetic node \u0f0b.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Punctuation", | |
| "sec_num": "3.3" | |
| }, | |
| { | |
| "text": "As one of the most widely-used punctuation in Tibetan, phonetic node \u0f0b is used to separate two syllables and vertical symbol \u0f0d to indicate boundary of two words, pause of words and sentences, and full stop of sentences. In some cases, there are exception in the usage of them. When a sentence is ended with the character \u0f42, the sentence ends without vertical symbol \u0f0d such as example (10). If the last letter of a sentence is \u0f44, the sentence ends with the combination of \u0f0b and \u0f0d as shown in example (11).", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Punctuation", | |
| "sec_num": "3.3" | |
| }, | |
| { | |
| "text": "Normally, the phonetic node \u0f0b is used similar with the white space in English such as example (13). In the situation of reduction raised by case, the phonetic node \u0f0b will be removed. For instance, in example (14) phonetic node \u0f0b of the syllable \u0f54\u0f0b is removed the case \u0f60\u0f72 is reduced.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Punctuation", | |
| "sec_num": "3.3" | |
| }, | |
| { | |
| "text": "Following the evaluation approach in early work (Soto et al., 2017; Fuentes et al., 2018) , SimpleNLG-TI is evaluated by comparing its output with a golden standard. When an output is completely identical with the golden standard, then we take SimpleNLG-TI passes this unit test.", | |
| "cite_spans": [ | |
| { | |
| "start": 48, | |
| "end": 67, | |
| "text": "(Soto et al., 2017;", | |
| "ref_id": "BIBREF15" | |
| }, | |
| { | |
| "start": 68, | |
| "end": 89, | |
| "text": "Fuentes et al., 2018)", | |
| "ref_id": "BIBREF4" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Evaluation", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "In this work, we take a test set as our golden standard, which is manually translated from the test set of SimpleNLG-EN. SimpleNLG-TI has successfully passed all 84 unit tests, which cover Tibetan linguistic features previous described in this paper. The result shows that this work can be used as a practical tool for NLG in Tibetan.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Evaluation", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "In this work, we propose SimpleNLG-TI (a Surface Realiser) for Tibetan generation, which is based on the framework of SimpleNLG-EN. We investigate main linguistic features (morphology and syntactic characteristics) and apply them into the development of SimpleNLG-TI. The realisation results of SimpleNLG-TI show that it is capable of generating well-formed text and is feasible to be deployed in Tibetan generation systems.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Conclusion", | |
| "sec_num": "5" | |
| }, | |
| { | |
| "text": "In the future, we plan to improve the performance of SimpleNLG-TI by taking more features into account, for instance, other types of phrases like adjective phrases and complex words and phrases. Moveover, we would like to test SimpleNLG-TI with practical Tibetan generation systems. The full Python package of SimpleNLG-TI will be publicly released on Github 2 .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Conclusion", | |
| "sec_num": "5" | |
| }, | |
| { | |
| "text": "We referred to SimpleNLG v4.4.8 rewrite by Bjascob https://github.com/bjascob/pySimpleNLG.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "https://github.com/ZWKZ/SimpleNLG-TI", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| } | |
| ], | |
| "back_matter": [ | |
| { | |
| "text": "The authors of this paper received support from Qinghai Natural Science Foundation un-derGrant 2016-ZJ-931Q, Qinghai Major RD Transformation Foundation under Grant 2019-GX-162, and National Natural Foundation under Grant 61862055, which is gratefully acknowledged.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Acknowledgments", | |
| "sec_num": null | |
| } | |
| ], | |
| "bib_entries": { | |
| "BIBREF0": { | |
| "ref_id": "b0", | |
| "title": "Enabling technology for multilingual natural language generation: the kpml development environment", | |
| "authors": [ | |
| { | |
| "first": "A", | |
| "middle": [], | |
| "last": "John", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "", | |
| "middle": [], | |
| "last": "Bateman", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1997, | |
| "venue": "Natural Language Engineering", | |
| "volume": "3", | |
| "issue": "1", | |
| "pages": "15--55", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "John A Bateman. 1997. Enabling technology for multilingual natural language generation: the kpml development environment. Natural Lan- guage Engineering, 3(1):15-55.", | |
| "links": null | |
| }, | |
| "BIBREF1": { | |
| "ref_id": "b1", | |
| "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 Florian Matthes. 2019. Simplenlg-de: Adapting simplenlg 4 to german. In Proceedings of the 12th International Conference on Natural Lan- guage Generation, pages 415-420.", | |
| "links": null | |
| }, | |
| "BIBREF3": { | |
| "ref_id": "b3", | |
| "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 International Conference on Natural Language Generation, pages 57-66.", | |
| "links": null | |
| }, | |
| "BIBREF4": { | |
| "ref_id": "b4", | |
| "title": "Adapting simplenlg to galician language", | |
| "authors": [ | |
| { | |
| "first": "Andrea Cascallar", | |
| "middle": [], | |
| "last": "Fuentes", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Alejandro", | |
| "middle": [ | |
| "Ramos" | |
| ], | |
| "last": "Soto", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Alberto", | |
| "middle": [ | |
| "Bugarin" | |
| ], | |
| "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 Bugarin Diz. 2018. Adapting sim- plenlg to galician language. Proceedings of the 11th International Conference on Natural Lan- guage Generation, pages 67-72.", | |
| "links": null | |
| }, | |
| "BIBREF5": { | |
| "ref_id": "b5", | |
| "title": "Simplenlg: A realisation engine for practical applications", | |
| "authors": [ | |
| { | |
| "first": "Albert", | |
| "middle": [], | |
| "last": "Gatt", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Ehud", | |
| "middle": [], | |
| "last": "Reiter", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2009, | |
| "venue": "Proceedings of the 12th European Workshop on Natural Language Generation", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "90--93", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Albert Gatt and Ehud Reiter. 2009. Simplenlg: A realisation engine for practical applications. In Proceedings of the 12th European Workshop on Natural Language Generation (ENLG 2009), pages 90-93.", | |
| "links": null | |
| }, | |
| "BIBREF6": { | |
| "ref_id": "b6", | |
| "title": "The Grammatical Functions of the Tibetan Adverbs", | |
| "authors": [ | |
| { | |
| "first": "", | |
| "middle": [], | |
| "last": "Dao Jieben", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2012, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Dao Jieben. 2012. The Grammatical Functions of the Tibetan Adverbs. Ph.D. thesis, Northwest University for Nationalities.", | |
| "links": null | |
| }, | |
| "BIBREF7": { | |
| "ref_id": "b7", | |
| "title": "Practical tibetan grammar", | |
| "authors": [ | |
| { | |
| "first": "Gesang", | |
| "middle": [], | |
| "last": "Jumian", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1987, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Gesang Jumian. 1987. Practical tibetan grammar. Sichuan Minorities Press.", | |
| "links": null | |
| }, | |
| "BIBREF8": { | |
| "ref_id": "b8", | |
| "title": "A framework for multi-document abstractive summarization based on semantic role labelling", | |
| "authors": [ | |
| { | |
| "first": "Atif", | |
| "middle": [], | |
| "last": "Khan", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Naomie", | |
| "middle": [], | |
| "last": "Salim", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Yogan Jaya", | |
| "middle": [], | |
| "last": "Kumar", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2015, | |
| "venue": "Applied Soft Computing", | |
| "volume": "30", | |
| "issue": "", | |
| "pages": "737--747", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Atif Khan, Naomie Salim, and Yogan Jaya Kumar. 2015. A framework for multi-document abstrac- tive summarization based on semantic role la- belling. Applied Soft Computing, 30:737-747.", | |
| "links": null | |
| }, | |
| "BIBREF9": { | |
| "ref_id": "b9", | |
| "title": "Sandor Csoma de K\u00f6r\u00f6s. 1834. A grammar of the Tibetan language", | |
| "authors": [], | |
| "year": null, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Sandor Csoma de K\u00f6r\u00f6s. 1834. A grammar of the Tibetan language. Baptist Mission Press.", | |
| "links": null | |
| }, | |
| "BIBREF10": { | |
| "ref_id": "b10", | |
| "title": "Recognition of tibetan linking verb and existential verb", | |
| "authors": [ | |
| { | |
| "first": "Lin", | |
| "middle": [], | |
| "last": "Li", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Congjun", | |
| "middle": [], | |
| "last": "Long", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2012, | |
| "venue": "Journal of Chinese Information Processing", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Lin Li and Congjun Long. 2012. Recognition of ti- betan linking verb and existential verb. In Jour- nal of Chinese Information Processing.", | |
| "links": null | |
| }, | |
| "BIBREF11": { | |
| "ref_id": "b11", | |
| "title": "Building large scale text corpus for tibetan natural language processing by extracting text from web", | |
| "authors": [ | |
| { | |
| "first": "Huidan", | |
| "middle": [], | |
| "last": "Liu", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Minghua", | |
| "middle": [], | |
| "last": "Nuo", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Jian", | |
| "middle": [], | |
| "last": "Wu", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Yeping", | |
| "middle": [], | |
| "last": "He", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2012, | |
| "venue": "24th International Conference on Computational Linguistics", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Huidan Liu, Minghua Nuo, Jian Wu, and Yeping He. 2012. Building large scale text corpus for ti- betan natural language processing by extracting text from web. In 24th International Conference on Computational Linguistics, page 11.", | |
| "links": null | |
| }, | |
| "BIBREF12": { | |
| "ref_id": "b12", | |
| "title": "Yag: A template-based generator for real-time systems", | |
| "authors": [ | |
| { | |
| "first": "Songsak", | |
| "middle": [], | |
| "last": "Susan W Mcroy", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "", | |
| "middle": [], | |
| "last": "Channarukul", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "S", | |
| "middle": [], | |
| "last": "Syed", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "", | |
| "middle": [], | |
| "last": "Ali", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2000, | |
| "venue": "Proceedings of the first international conference on Natural language generation", | |
| "volume": "14", | |
| "issue": "", | |
| "pages": "264--267", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Susan W McRoy, Songsak Channarukul, and Syed S Ali. 2000. Yag: A template-based gen- erator for real-time systems. In Proceedings of the first international conference on Natural lan- guage generation-Volume 14, pages 264-267. As- sociation for Computational Linguistics.", | |
| "links": null | |
| }, | |
| "BIBREF13": { | |
| "ref_id": "b13", | |
| "title": "Building applied natural language generation systems", | |
| "authors": [ | |
| { | |
| "first": "Ehud", | |
| "middle": [], | |
| "last": "Reiter", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Robert", | |
| "middle": [], | |
| "last": "Dale", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1997, | |
| "venue": "Natural Language Engineering", | |
| "volume": "3", | |
| "issue": "1", | |
| "pages": "57--87", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Ehud Reiter and Robert Dale. 1997. Building ap- plied natural language generation systems. Nat- ural Language Engineering, 3(1):57-87.", | |
| "links": null | |
| }, | |
| "BIBREF14": { | |
| "ref_id": "b14", | |
| "title": "Generation of formal and informal sentences", | |
| "authors": [ | |
| { | |
| "first": "Abu", | |
| "middle": [], | |
| "last": "Fadi", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Diana", | |
| "middle": [], | |
| "last": "Sheikha", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "", | |
| "middle": [], | |
| "last": "Inkpen", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2011, | |
| "venue": "Proceedings of the 13th European Workshop on Natural Language Generation", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "187--193", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Fadi Abu Sheikha and Diana Inkpen. 2011. Gener- ation of formal and informal sentences. In Pro- ceedings of the 13th European Workshop on Nat- ural Language Generation, pages 187-193. Asso- ciation for Computational Linguistics.", | |
| "links": null | |
| }, | |
| "BIBREF15": { | |
| "ref_id": "b15", | |
| "title": "Adapting simplenlg to spanish", | |
| "authors": [ | |
| { | |
| "first": "Alejandro", | |
| "middle": [ | |
| "Ramos" | |
| ], | |
| "last": "Soto", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Julio", | |
| "middle": [ | |
| "Janeiro" | |
| ], | |
| "last": "Gallardo", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Alberto", | |
| "middle": [ | |
| "Bugarin" | |
| ], | |
| "last": "Diz", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2017, | |
| "venue": "Proceedings of the 10th International Conference on Natural Language Generation", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Alejandro Ramos Soto, Julio Janeiro Gallardo, and Alberto Bugarin Diz. 2017. Adapting simplenlg to spanish. Proceedings of the 10th International Conference on Natural Language Generation.", | |
| "links": null | |
| }, | |
| "BIBREF16": { | |
| "ref_id": "b16", | |
| "title": "Qugan: Quasi generative adversarial network for tibetan question answering corpus generation", | |
| "authors": [ | |
| { | |
| "first": "Yuan", | |
| "middle": [], | |
| "last": "Sun", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Chaofan", | |
| "middle": [], | |
| "last": "Chen", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Tianci", | |
| "middle": [], | |
| "last": "Xia", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Xiaobing", | |
| "middle": [], | |
| "last": "Zhao", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2019, | |
| "venue": "IEEE Access", | |
| "volume": "7", | |
| "issue": "", | |
| "pages": "116247--116255", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Yuan Sun, Chaofan Chen, Tianci Xia, and Xiaob- ing Zhao. 2019. Qugan: Quasi generative ad- versarial network for tibetan question answer- ing corpus generation. IEEE Access, 7:116247- 116255.", | |
| "links": null | |
| }, | |
| "BIBREF17": { | |
| "ref_id": "b17", | |
| "title": "Adapting simplenlg for bilingual english-french realisation", | |
| "authors": [ | |
| { | |
| "first": "Pierre-Luc", | |
| "middle": [], | |
| "last": "Vaudry", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Guy", | |
| "middle": [], | |
| "last": "Lapalme", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2013, | |
| "venue": "Proceedings of the 14th European Workshop on Natural Language Generation", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "183--187", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Pierre-Luc Vaudry and Guy Lapalme. 2013. Adapting simplenlg for bilingual english-french realisation. In Proceedings of the 14th Euro- pean Workshop on Natural Language Genera- tion, pages 183-187.", | |
| "links": null | |
| }, | |
| "BIBREF18": { | |
| "ref_id": "b18", | |
| "title": "Ccg chart realization from disjunctive inputs", | |
| "authors": [ | |
| { | |
| "first": "Michael", | |
| "middle": [], | |
| "last": "White", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2006, | |
| "venue": "Proceedings of the Fourth International Natural Language Generation Conference", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "12--19", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Michael White. 2006. Ccg chart realization from disjunctive inputs. In Proceedings of the Fourth International Natural Language Genera- tion Conference, pages 12-19.", | |
| "links": null | |
| }, | |
| "BIBREF19": { | |
| "ref_id": "b19", | |
| "title": "Design and research of tibetan encyclopedia knowledge question answering system", | |
| "authors": [ | |
| { | |
| "first": "Chen", | |
| "middle": [], | |
| "last": "Xiaoying", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2017, | |
| "venue": "Intelligent Computer and Application", | |
| "volume": "7", | |
| "issue": "4", | |
| "pages": "48--50", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Chen Xiaoying. 2017. Design and research of ti- betan encyclopedia knowledge question answer- ing system. Intelligent Computer and Applica- tion, 7(4):48-50.", | |
| "links": null | |
| }, | |
| "BIBREF20": { | |
| "ref_id": "b20", | |
| "title": "Research on tibetan-chinese neural machine translation", | |
| "authors": [ | |
| { | |
| "first": "Li", | |
| "middle": [], | |
| "last": "Yachao", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2017, | |
| "venue": "Journal of Chinese Information Processing", | |
| "volume": "31", | |
| "issue": "6", | |
| "pages": "103--109", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Li Yachao. 2017. Research on tibetan-chinese neu- ral machine translation. Journal of Chinese In- formation Processing, 31(6):103-109.", | |
| "links": null | |
| } | |
| }, | |
| "ref_entries": { | |
| "FIGREF0": { | |
| "num": null, | |
| "text": "13) \u0f58\u0f7a\u0f0b\u0f4f\u0f7c\u0f42\u0f0b + \u0f58\u0f5b\u0f7a\u0f66\u0f0b\u0f54\u0f0b = \u0f58\u0f7a\u0f0b\u0f4f\u0f7c\u0f42\u0f0b\u0f58\u0f5b\u0f7a\u0f66\u0f0b\u0f54\u0f0b flower + beautiful = flower beautiful \u0f62\u0f9f\u0f0b + \u0f53\u0f7c\u0f62\u0f0b + \u0f63\u0f74\u0f42\u0f0b = \u0f62\u0f9f\u0f0b\u0f53\u0f7c\u0f62\u0f0b\u0f63\u0f74\u0f42\u0f0b horse + yak + sheep = horse yak sheep (14) \u0f58\u0f5b\u0f7a\u0f66\u0f0b\u0f54\u0f0b + \u0f60\u0f72\u0f0b + \u0f58\u0f7a\u0f0b\u0f4f\u0f7c\u0f42\u0f0b = \u0f58\u0f5b\u0f7a\u0f66\u0f0b\u0f54\u0f60\u0f72\u0f0b\u0f58\u0f7a\u0f0b\u0f4f\u0f42\u0f0b beauty + case + flower = beautiful flower (15) \u0f64\u0f72\u0f53\u0f0b\u0f4f\u0f74\u0f0b + \u0f58\u0f5b\u0f7a\u0f66\u0f0b = \u0f64\u0f72\u0f53\u0f0b\u0f4f\u0f74\u0f0b\u0f58\u0f5b\u0f7a\u0f66\u0f0b very + beautiful = very beautiful", | |
| "uris": null, | |
| "type_str": "figure" | |
| }, | |
| "TABREF0": { | |
| "num": null, | |
| "html": null, | |
| "text": "An instance of realising a sentence with SimpleNLG-TI.", | |
| "type_str": "table", | |
| "content": "<table/>" | |
| }, | |
| "TABREF1": { | |
| "num": null, | |
| "html": null, | |
| "text": "Tibetan verbs have three forms of inflection, that is tense, imperative and the change of the voices. Clearly rules of morphology changes are not founded yet, therefore SimpleNLG-TI has adopted a simplified model for verb realising. example (6) is the three forms of verb \u0f5f eat.", | |
| "type_str": "table", | |
| "content": "<table><tr><td colspan=\"2\">3.1.2 Verb Tense</td><td/><td/></tr><tr><td/><td colspan=\"3\">example (7) is a special case</td></tr><tr><td colspan=\"4\">whose three inflection are the same, which</td></tr><tr><td colspan=\"4\">is not uncommon in Tibetan. The lexicon</td></tr><tr><td colspan=\"4\">of SimpleNLG-TI contains more than 1,500</td></tr><tr><td colspan=\"4\">monosyllabic verbs for now, which includes the</td></tr><tr><td colspan=\"3\">above inflection information.</td><td/></tr><tr><td>(6) \u0f5f</td><td>\u0f56\u0f5f\u0f66</td><td>\u0f56\u0f5f\u0f60</td><td>\u0f5f\u0f7c</td></tr><tr><td>eat</td><td colspan=\"3\">ate will eat do eat</td></tr><tr><td colspan=\"4\">present past future imperative</td></tr><tr><td>(7) \u0f62\u0f7a\u0f42</td><td>\u0f62\u0f7a\u0f42</td><td>\u0f62\u0f7a\u0f42</td><td>\u0f62\u0f7a\u0f42</td></tr><tr><td colspan=\"4\">touch touched will touch do touch</td></tr><tr><td colspan=\"2\">present past</td><td>future</td><td>imperative</td></tr><tr><td/><td/><td/><td>) \u0f58\u0f51\u0f74\u0f53\u0f0b \u0f45\u0f7c\u0f42(table)</td><td>\u0f58\u0f51\u0f74\u0f53\u0f0b \u0f45\u0f7c\u0f42\u0f0b\u0f51\u0f42(tables)</td></tr><tr><td/><td/><td/><td>(4) \u0f44(I )</td><td>\u0f44\u0f0b\u0f5a\u0f7c(we)</td></tr><tr><td/><td/><td/><td>(5) \u0f41\u0f7c(he)</td><td>\u0f41\u0f7c\u0f0b\u0f45\u0f42(they)</td></tr></table>" | |
| }, | |
| "TABREF2": { | |
| "num": null, | |
| "html": null, | |
| "text": "\u0f58\u0f7a\u0f51 and \u0f58\u0f72\u0f53 is negation of \u0f61\u0f7c\u0f51 and \u0f61\u0f72\u0f53.", | |
| "type_str": "table", | |
| "content": "<table><tr><td>(10) \uf5d4\u0f7c\u0f56\u0f0b\u0f42\u0fb2\u0fad\u0f0b\u0f60\u0f42\u0fb2\u0f7c</td><td>\uf5d4\u0f7c\u0f56\u0f0b\u0f42\u0fb2\u0fad\u0f0b\u0f58\u0f72\u0f0b\u0f60\u0f42\u0fb2\u0f7c</td></tr><tr><td>go to school \u0f58\u0f0b\u0f60\u0f7c\u0f44\u0f66\u0f0b\u0f54\u0f62\u0f0b\uf5d4\u0f7c\u0f56\u0f0b\u0f42\u0fb2\u0fad\u0f0b\u0f60\u0f42\u0fb2\u0f7c</td><td>don t go to school \u0f58\u0f0b\u0f60\u0f7c\u0f44\u0f66\u0f0b\u0f54\u0f62\u0f0b\uf5d4\u0f7c\u0f56\u0f0b\u0f42\u0fb2\u0fad\u0f0b\u0f63\u0f0b\u0f58\u0f72\u0f0b\u0f60\u0f42\u0fb2\u0f7c</td></tr><tr><td>will go to school</td><td>will not go to school</td></tr><tr><td>(11) \uf5d4\u0f7c\u0f56\u0f0b\u0f42\u0fb2\u0fad\u0f62\u0f0b\u0f66\u0f7c\u0f44\u0f0b\u0f0d</td><td>\uf5d4\u0f7c\u0f56\u0f0b\u0f42\u0fb2\u0fad\u0f62\u0f0b\u0f58\u0f0b\u0f66\u0f7c\u0f44\u0f0b\u0f0d</td></tr><tr><td>went to school</td><td>didn't go to school</td></tr><tr><td>3. (12) \u0f44\u0f0b\u0f53\u0f72\u0f0b\u0f51\u0f42\u0f7a\u0f0b\u0f62\u0f92\u0f53\u0f0b\u0f61\u0f72\u0f53\u0f0d</td><td>\u0f44\u0f0b\u0f53\u0f72\u0f0b\u0f51\u0f42\u0f7a\u0f0b\u0f62\u0f92\u0f53\u0f0b\u0f58\u0f72\u0f53\u0f0d</td></tr><tr><td>I'm a teacher. \u0f44\u0f0b\u0f63\u0f0b\u0f51\u0f54\u0f7a\u0f0b\u0f46\u0f0b\u0f61\u0f7c\u0f51\u0f0d</td><td>I'm not a teacher. \u0f44\u0f0b\u0f63\u0f0b\u0f51\u0f54\u0f7a\u0f0b\u0f46\u0f0b\u0f58\u0f7a\u0f51\u0f0d</td></tr><tr><td>I have book</td><td>I don t have book.</td></tr></table>" | |
| } | |
| } | |
| } | |
| } |