{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/reporting-score-distributions-makes-a","title":"Reporting Score Distributions Makes a Difference: Performance Study of LSTM-networks for Sequence Tagging","arxiv_id":"1707.09861","date":"2017-07-31","proceeding":"EMNLP 2017 9","authors":["Nils Reimers","Iryna Gurevych"],"abstract":"In this paper we show that reporting a single performance score is\ninsufficient to compare non-deterministic approaches. We demonstrate for common\nsequence tagging tasks that the seed value for the random number generator can\nresult in statistically significant (p < 10^-4) differences for\nstate-of-the-art systems. For two recent systems for NER, we observe an\nabsolute difference of one percentage point F1-score depending on the selected\nseed value, making these systems perceived either as state-of-the-art or\nmediocre. Instead of publishing and reporting single performance scores, we\npropose to compare score distributions based on multiple executions. Based on\nthe evaluation of 50.000 LSTM-networks for five sequence tagging tasks, we\npresent network architectures that produce both superior performance as well as\nare more stable with respect to the remaining hyperparameters.","url_abs":"http://arxiv.org/abs/1707.09861v1","url_pdf":"http://arxiv.org/pdf/1707.09861v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"reporting-score-distributions-makes-a","repo_url":"https://github.com/UKPLab/emnlp2017-bilstm-cnn-crf","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"reporting-score-distributions-makes-a","repo_url":"https://github.com/UKPLab/elmo-bilstm-cnn-crf","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"reporting-score-distributions-makes-a","repo_url":"https://github.com/tommasoc80/Event_detection_CLiC-it2018","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"reporting-score-distributions-makes-a","repo_url":"https://github.com/vietnlp/etnlp","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"reporting-score-distributions-makes-a","repo_url":"https://github.com/vinhdv1628/etnlp","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"named-entity-recognition-ner","task_name":"Named Entity Recognition (NER)"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1707.09861","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}