| { | |
| "paper_id": "M93-1008", | |
| "header": { | |
| "generated_with": "S2ORC 1.0.0", | |
| "date_generated": "2023-01-19T03:14:20.340745Z" | |
| }, | |
| "title": "THE STATISTICAL SIGNIFICANCE OF THE MUC-5 RESULT S", | |
| "authors": [ | |
| { | |
| "first": "Nancy", | |
| "middle": [], | |
| "last": "Chinchor", | |
| "suffix": "", | |
| "affiliation": {}, | |
| "email": "chinchor@gso.saic.com" | |
| } | |
| ], | |
| "year": "", | |
| "venue": null, | |
| "identifiers": {}, | |
| "abstract": "", | |
| "pdf_parse": { | |
| "paper_id": "M93-1008", | |
| "_pdf_hash": "", | |
| "abstract": [], | |
| "body_text": [ | |
| { | |
| "text": "The statistical significance of the results of the MUC-5 evaluation is determined using a computer-intensiv e method of hypothesis testing known as approximate randomization . The exact method is described in detail in 111 an d [2] and has been used as the accepted statistical test for the MUC results since MUC-3 . The purpose of the statistica l testing is to determine whether the scores of the systems are different by chance or due to a significant difference i n the character of the systems .", | |
| "cite_spans": [ | |
| { | |
| "start": 228, | |
| "end": 231, | |
| "text": "[2]", | |
| "ref_id": null | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "INTRODUCTION", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Statistical significance results are reported here for the following metrics : Error per Response Fill, F-Mcasure with recall and precision weighted equally, and Richness-Normalized Error (minimum and maximum) . The systems are compared for the same domain and language and, thus, there are four figures for each metric : English Join t Ventures (EJV), Japanese Joint Ventures (JJV), English Microelectronics (EME), and Japanese Microelectronics (JME) . The format of the reporting is according to the groupings of the systems which are not significantly differen t from each other at the 0.01 level with a confidence of at least 99% . Systems which are not significantly different fro m each other are underscored on the same line . The systems are numbered to save space and the correspondence of th e number and system site are given below the significance results .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "STATISTICAL SIGNIFICANCE RESULT S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "It is interesting to note that the rankings of systems do not change when using the Error per Response Fil l metric or the F-Measure. The numerical rankings change slightly (numbers 6 and 7 in EJV reverse, and numbers 4 and 5 in JJV reverse), but those changes are not significant statistically because the two members in each of th e reversed pairs are both in the same significance grouping for both of the two metrics . It is also interesting to note tha t the Error per Response Fill metric distinguishes four more systems than the F-Measure over all domains and languages . The Richness-Normalized Error metric distinguishes far fewer systems statistically than the Error pe r Response Fill metric with 29 systems distinguished by Richness-Normalized Error as opposed to 55 by Error per Response Fill for EJV alone . Both the minimum and maximum Richness-Normalized Error metrics produce the sam e rankings and statistical results so are conflated here . The statistical groupings of systems for Richness-Normalized Error are so large and so numerous that systems cannot be distinguished well enough to reflect their perceived differences in performance . It is believed that this is due to the fact that the Richness-Normalized Error metric ignores th e amount of spurious data generated by a system and that the amount and kind of spurious data generated impacts th e perception of how well the system is doing in an operational setting .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "STATISTICAL SIGNIFICANCE RESULT S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The approximate randomization method has been used to determine the statistical significance of the rankings of systems for MUC-5 . It is also useful for reflecting on the relative merits of the evaluation metrics . The statistical results show that the Error per Response Fill metric is the most sensitive metric of the three in terms o f distinguishing systems. However, no statistically significant changes in ranking occur when F-Measure is used . The Richness-Normalized Error metric distinguishes far fewer systems than either of the other metrics .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "CONCLUSIONS", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "English Joint Ventures -Error per Response Fill 1 2 3 4 5 6 7 8 9 10 11 12 13 1) GE/CMU 2) BBN 3) SRI 4) UMASS/HU 5) PMAX 6) USUSSEX 7) NMSUIBR 8) NYU 9) SRA 10) PRC 11) USC 12) MITRE 13)TRW panese Microelectronics -Richness-Normalized Error 1 2 3 4 5", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "CONCLUSIONS", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "EQUATION", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [ | |
| { | |
| "start": 0, | |
| "end": 8, | |
| "text": "EQUATION", | |
| "ref_id": "EQREF", | |
| "raw_str": "I) GE/CMU-OPT 2) GFJCMU 3) NMSU/BR 4) SRA 5)", | |
| "eq_num": "SRI" | |
| } | |
| ], | |
| "section": "CONCLUSIONS", | |
| "sec_num": null | |
| } | |
| ], | |
| "back_matter": [ | |
| { | |
| "text": "English Microelectronics -F-Measure (P&R ) 1 2 3 4 5 6 7English Joint Ventures -Richness-Normalized Erro r 1 2 3 4 5 6 7 8 9 10 11 12 132) SRI 3) NYU 4) UMASS/H U 9) MITRE I0) PRC 1 I) PMAX 5) GE/CMU 6) USUSSEX 7) NMSU/BR 12) USC 13) TRW", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "annex", | |
| "sec_num": null | |
| } | |
| ], | |
| "bib_entries": {}, | |
| "ref_entries": { | |
| "FIGREF0": { | |
| "type_str": "figure", | |
| "uris": null, | |
| "num": null, | |
| "text": "Chinchor, N ., L . Hirschman, and D . Lewis (1993) \"Evaluating Message Understanding Systems : An Analysi s of the Third Message Understanding Conference (MUC-3)\" Computational Linguistics 19(3) .[2] Chinchor, N. (1992) . \"The Statistical Significance of the MUC-4 Results\" Proceedings of the Fourth Message Understanding Conference (MUC-4) . Morgan Kaufmann, Publishers . San Mateo, CA ." | |
| } | |
| } | |
| } | |
| } |