Pub Date : 2021-07-24DOI: 10.18653/v1/2021.bionlp-1.13
Vladimir Araujo, Andrés Carvallo, C. Aspillaga, C. Thorne, Denis Parra
The success of pretrained word embeddings has motivated their use in the biomedical domain, with contextualized embeddings yielding remarkable results in several biomedical NLP tasks. However, there is a lack of research on quantifying their behavior under severe “stress” scenarios. In this work, we systematically evaluate three language models with adversarial examples – automatically constructed tests that allow us to examine how robust the models are. We propose two types of stress scenarios focused on the biomedical named entity recognition (NER) task, one inspired by spelling errors and another based on the use of synonyms for medical terms. Our experiments with three benchmarks show that the performance of the original models decreases considerably, in addition to revealing their weaknesses and strengths. Finally, we show that adversarial training causes the models to improve their robustness and even to exceed the original performance in some cases.
{"title":"Stress Test Evaluation of Biomedical Word Embeddings","authors":"Vladimir Araujo, Andrés Carvallo, C. Aspillaga, C. Thorne, Denis Parra","doi":"10.18653/v1/2021.bionlp-1.13","DOIUrl":"https://doi.org/10.18653/v1/2021.bionlp-1.13","url":null,"abstract":"The success of pretrained word embeddings has motivated their use in the biomedical domain, with contextualized embeddings yielding remarkable results in several biomedical NLP tasks. However, there is a lack of research on quantifying their behavior under severe “stress” scenarios. In this work, we systematically evaluate three language models with adversarial examples – automatically constructed tests that allow us to examine how robust the models are. We propose two types of stress scenarios focused on the biomedical named entity recognition (NER) task, one inspired by spelling errors and another based on the use of synonyms for medical terms. Our experiments with three benchmarks show that the performance of the original models decreases considerably, in addition to revealing their weaknesses and strengths. Finally, we show that adversarial training causes the models to improve their robustness and even to exceed the original performance in some cases.","PeriodicalId":200974,"journal":{"name":"Workshop on Biomedical Natural Language Processing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129324049","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-06-01DOI: 10.18653/v1/2021.bionlp-1.27
Lana Yeganova, Won Kim, Donald C. Comeau, W. Wilbur, Zhiyong Lu
With the growing availability of full-text articles, integrating abstracts and full texts of documents into a unified representation is essential for comprehensive search of scientific literature. However, previous studies have shown that naïvely merging abstracts with full texts of articles does not consistently yield better performance. Balancing the contribution of query terms appearing in the abstract and in sections of different importance in full text articles remains a challenge both with traditional bag-of-words IR approaches and for neural retrieval methods. In this work we establish the connection between the BM25 score of a query term appearing in a section of a full text document and the probability of that document being clicked or identified as relevant. Probability is computed using Pool Adjacent Violators (PAV), an isotonic regression algorithm, providing a maximum likelihood estimate based on the observed data. Using this probabilistic transformation of BM25 scores we show an improved performance on the PubMed Click dataset developed and presented in this study, as well as the 2007 TREC Genomics collection.
{"title":"Measuring the relative importance of full text sections for information retrieval from scientific literature.","authors":"Lana Yeganova, Won Kim, Donald C. Comeau, W. Wilbur, Zhiyong Lu","doi":"10.18653/v1/2021.bionlp-1.27","DOIUrl":"https://doi.org/10.18653/v1/2021.bionlp-1.27","url":null,"abstract":"With the growing availability of full-text articles, integrating abstracts and full texts of documents into a unified representation is essential for comprehensive search of scientific literature. However, previous studies have shown that naïvely merging abstracts with full texts of articles does not consistently yield better performance. Balancing the contribution of query terms appearing in the abstract and in sections of different importance in full text articles remains a challenge both with traditional bag-of-words IR approaches and for neural retrieval methods. In this work we establish the connection between the BM25 score of a query term appearing in a section of a full text document and the probability of that document being clicked or identified as relevant. Probability is computed using Pool Adjacent Violators (PAV), an isotonic regression algorithm, providing a maximum likelihood estimate based on the observed data. Using this probabilistic transformation of BM25 scores we show an improved performance on the PubMed Click dataset developed and presented in this study, as well as the 2007 TREC Genomics collection.","PeriodicalId":200974,"journal":{"name":"Workshop on Biomedical Natural Language Processing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130571536","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-30DOI: 10.18653/v1/2021.bionlp-1.19
Christoph Müller, Sucheta Ghosh, Ulrike Wittig, Maja Rey
We describe a simple procedure for the automatic creation of word-level alignments between printed documents and their respective full-text versions. The procedure is unsupervised, uses standard, off-the-shelf components only, and reaches an F-score of 85.01 in the basic setup and up to 86.63 when using pre- and post-processing. Potential areas of application are manual database curation (incl. document triage) and biomedical expression OCR.
{"title":"Word-Level Alignment of Paper Documents with their Electronic Full-Text Counterparts","authors":"Christoph Müller, Sucheta Ghosh, Ulrike Wittig, Maja Rey","doi":"10.18653/v1/2021.bionlp-1.19","DOIUrl":"https://doi.org/10.18653/v1/2021.bionlp-1.19","url":null,"abstract":"We describe a simple procedure for the automatic creation of word-level alignments between printed documents and their respective full-text versions. The procedure is unsupervised, uses standard, off-the-shelf components only, and reaches an F-score of 85.01 in the basic setup and up to 86.63 when using pre- and post-processing. Potential areas of application are manual database curation (incl. document triage) and biomedical expression OCR.","PeriodicalId":200974,"journal":{"name":"Workshop on Biomedical Natural Language Processing","volume":"96 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127084192","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-28DOI: 10.18653/v1/2021.bionlp-1.1
P. Su, Yifan Peng, K. Vijay-Shanker
Contrastive learning has been used to learn a high-quality representation of the image in computer vision. However, contrastive learning is not widely utilized in natural language processing due to the lack of a general method of data augmentation for text data. In this work, we explore the method of employing contrastive learning to improve the text representation from the BERT model for relation extraction. The key knob of our framework is a unique contrastive pre-training step tailored for the relation extraction tasks by seamlessly integrating linguistic knowledge into the data augmentation. Furthermore, we investigate how large-scale data constructed from the external knowledge bases can enhance the generality of contrastive pre-training of BERT. The experimental results on three relation extraction benchmark datasets demonstrate that our method can improve the BERT model representation and achieve state-of-the-art performance. In addition, we explore the interpretability of models by showing that BERT with contrastive pre-training relies more on rationales for prediction. Our code and data are publicly available at: https://github.com/AnonymousForNow.
{"title":"Improving BERT Model Using Contrastive Learning for Biomedical Relation Extraction","authors":"P. Su, Yifan Peng, K. Vijay-Shanker","doi":"10.18653/v1/2021.bionlp-1.1","DOIUrl":"https://doi.org/10.18653/v1/2021.bionlp-1.1","url":null,"abstract":"Contrastive learning has been used to learn a high-quality representation of the image in computer vision. However, contrastive learning is not widely utilized in natural language processing due to the lack of a general method of data augmentation for text data. In this work, we explore the method of employing contrastive learning to improve the text representation from the BERT model for relation extraction. The key knob of our framework is a unique contrastive pre-training step tailored for the relation extraction tasks by seamlessly integrating linguistic knowledge into the data augmentation. Furthermore, we investigate how large-scale data constructed from the external knowledge bases can enhance the generality of contrastive pre-training of BERT. The experimental results on three relation extraction benchmark datasets demonstrate that our method can improve the BERT model representation and achieve state-of-the-art performance. In addition, we explore the interpretability of models by showing that BERT with contrastive pre-training relies more on rationales for prediction. Our code and data are publicly available at: https://github.com/AnonymousForNow.","PeriodicalId":200974,"journal":{"name":"Workshop on Biomedical Natural Language Processing","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133745703","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-23DOI: 10.18653/v1/2021.bionlp-1.15
Amelie Wuhrl, Roman Klinger
Social media contains unfiltered and unique information, which is potentially of great value, but, in the case of misinformation, can also do great harm. With regards to biomedical topics, false information can be particularly dangerous. Methods of automatic fact-checking and fake news detection address this problem, but have not been applied to the biomedical domain in social media yet. We aim to fill this research gap and annotate a corpus of 1200 tweets for implicit and explicit biomedical claims (the latter also with span annotations for the claim phrase). With this corpus, which we sample to be related to COVID-19, measles, cystic fibrosis, and depression, we develop baseline models which detect tweets that contain a claim automatically. Our analyses reveal that biomedical tweets are densely populated with claims (45 % in a corpus sampled to contain 1200 tweets focused on the domains mentioned above). Baseline classification experiments with embedding-based classifiers and BERT-based transfer learning demonstrate that the detection is challenging, however, shows acceptable performance for the identification of explicit expressions of claims. Implicit claim tweets are more challenging to detect.
{"title":"Claim Detection in Biomedical Twitter Posts","authors":"Amelie Wuhrl, Roman Klinger","doi":"10.18653/v1/2021.bionlp-1.15","DOIUrl":"https://doi.org/10.18653/v1/2021.bionlp-1.15","url":null,"abstract":"Social media contains unfiltered and unique information, which is potentially of great value, but, in the case of misinformation, can also do great harm. With regards to biomedical topics, false information can be particularly dangerous. Methods of automatic fact-checking and fake news detection address this problem, but have not been applied to the biomedical domain in social media yet. We aim to fill this research gap and annotate a corpus of 1200 tweets for implicit and explicit biomedical claims (the latter also with span annotations for the claim phrase). With this corpus, which we sample to be related to COVID-19, measles, cystic fibrosis, and depression, we develop baseline models which detect tweets that contain a claim automatically. Our analyses reveal that biomedical tweets are densely populated with claims (45 % in a corpus sampled to contain 1200 tweets focused on the domains mentioned above). Baseline classification experiments with embedding-based classifiers and BERT-based transfer learning demonstrate that the detection is challenging, however, shows acceptable performance for the identification of explicit expressions of claims. Implicit claim tweets are more challenging to detect.","PeriodicalId":200974,"journal":{"name":"Workshop on Biomedical Natural Language Processing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133826230","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pretrained language models have shown success in many natural language processing tasks. Many works explore to incorporate the knowledge into the language models. In the biomedical domain, experts have taken decades of effort on building large-scale knowledge bases. For example, UMLS contains millions of entities with their synonyms and defines hundreds of relations among entities. Leveraging this knowledge can benefit a variety of downstream tasks such as named entity recognition and relation extraction. To this end, we propose KeBioLM, a biomedical pretrained language model that explicitly leverages knowledge from the UMLS knowledge bases. Specifically, we extract entities from PubMed abstracts and link them to UMLS. We then train a knowledge-aware language model that firstly applies a text-only encoding layer to learn entity representation and then applies a text-entity fusion encoding to aggregate entity representation. In addition, we add two training objectives as entity detection and entity linking. Experiments on the named entity recognition and relation extraction tasks from the BLURB benchmark demonstrate the effectiveness of our approach. Further analysis on a collected probing dataset shows that our model has better ability to model medical knowledge.
{"title":"Improving Biomedical Pretrained Language Models with Knowledge","authors":"Zheng Yuan, Yijia Liu, Chuanqi Tan, Songfang Huang, Fei Huang","doi":"10.18653/v1/2021.bionlp-1.20","DOIUrl":"https://doi.org/10.18653/v1/2021.bionlp-1.20","url":null,"abstract":"Pretrained language models have shown success in many natural language processing tasks. Many works explore to incorporate the knowledge into the language models. In the biomedical domain, experts have taken decades of effort on building large-scale knowledge bases. For example, UMLS contains millions of entities with their synonyms and defines hundreds of relations among entities. Leveraging this knowledge can benefit a variety of downstream tasks such as named entity recognition and relation extraction. To this end, we propose KeBioLM, a biomedical pretrained language model that explicitly leverages knowledge from the UMLS knowledge bases. Specifically, we extract entities from PubMed abstracts and link them to UMLS. We then train a knowledge-aware language model that firstly applies a text-only encoding layer to learn entity representation and then applies a text-entity fusion encoding to aggregate entity representation. In addition, we add two training objectives as entity detection and entity linking. Experiments on the named entity recognition and relation extraction tasks from the BLURB benchmark demonstrate the effectiveness of our approach. Further analysis on a collected probing dataset shows that our model has better ability to model medical knowledge.","PeriodicalId":200974,"journal":{"name":"Workshop on Biomedical Natural Language Processing","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126108488","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-04-14DOI: 10.18653/v1/2021.bionlp-1.6
Damian Pascual, Sandro Luck, Roger Wattenhofer
Automatic ICD coding is the task of assigning codes from the International Classification of Diseases (ICD) to medical notes. These codes describe the state of the patient and have multiple applications, e.g., computer-assisted diagnosis or epidemiological studies. ICD coding is a challenging task due to the complexity and length of medical notes. Unlike the general trend in language processing, no transformer model has been reported to reach high performance on this task. Here, we investigate in detail ICD coding using PubMedBERT, a state-of-the-art transformer model for biomedical language understanding. We find that the difficulty of fine-tuning the model on long pieces of text is the main limitation for BERT-based models on ICD coding. We run extensive experiments and show that despite the gap with current state-of-the-art, pretrained transformers can reach competitive performance using relatively small portions of text. We point at better methods to aggregate information from long texts as the main need for improving BERT-based ICD coding.
{"title":"Towards BERT-based Automatic ICD Coding: Limitations and Opportunities","authors":"Damian Pascual, Sandro Luck, Roger Wattenhofer","doi":"10.18653/v1/2021.bionlp-1.6","DOIUrl":"https://doi.org/10.18653/v1/2021.bionlp-1.6","url":null,"abstract":"Automatic ICD coding is the task of assigning codes from the International Classification of Diseases (ICD) to medical notes. These codes describe the state of the patient and have multiple applications, e.g., computer-assisted diagnosis or epidemiological studies. ICD coding is a challenging task due to the complexity and length of medical notes. Unlike the general trend in language processing, no transformer model has been reported to reach high performance on this task. Here, we investigate in detail ICD coding using PubMedBERT, a state-of-the-art transformer model for biomedical language understanding. We find that the difficulty of fine-tuning the model on long pieces of text is the main limitation for BERT-based models on ICD coding. We run extensive experiments and show that despite the gap with current state-of-the-art, pretrained transformers can reach competitive performance using relatively small portions of text. We point at better methods to aggregate information from long texts as the main need for improving BERT-based ICD coding.","PeriodicalId":200974,"journal":{"name":"Workshop on Biomedical Natural Language Processing","volume":"232 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121297301","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-07-01DOI: 10.18653/v1/2020.bionlp-1.17
Brian Hur, Timothy Baldwin, Karin M. Verspoor, L. Hardefeldt, J. Gilkerson
Identifying the reasons for antibiotic administration in veterinary records is a critical component of understanding antimicrobial usage patterns. This informs antimicrobial stewardship programs designed to fight antimicrobial resistance, a major health crisis affecting both humans and animals in which veterinarians have an important role to play. We propose a document classification approach to determine the reason for administration of a given drug, with particular focus on domain adaptation from one drug to another, and instance selection to minimize annotation effort.
{"title":"Domain Adaptation and Instance Selection for Disease Syndrome Classification over Veterinary Clinical Notes","authors":"Brian Hur, Timothy Baldwin, Karin M. Verspoor, L. Hardefeldt, J. Gilkerson","doi":"10.18653/v1/2020.bionlp-1.17","DOIUrl":"https://doi.org/10.18653/v1/2020.bionlp-1.17","url":null,"abstract":"Identifying the reasons for antibiotic administration in veterinary records is a critical component of understanding antimicrobial usage patterns. This informs antimicrobial stewardship programs designed to fight antimicrobial resistance, a major health crisis affecting both humans and animals in which veterinarians have an important role to play. We propose a document classification approach to determine the reason for administration of a given drug, with particular focus on domain adaptation from one drug to another, and instance selection to minimize annotation effort.","PeriodicalId":200974,"journal":{"name":"Workshop on Biomedical Natural Language Processing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130705879","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-07-01DOI: 10.18653/v1/2020.bionlp-1.5
A. Koroleva, S. Kamath, P. Bossuyt, P. Paroubek
Improving the quality of medical research reporting is crucial to reduce avoidable waste in research and to improve the quality of health care. Despite various initiatives aiming at improving research reporting – guidelines, checklists, authoring aids, peer review procedures, etc. – overinterpretation of research results, also known as spin, is still a serious issue in research reporting. In this paper, we propose a Natural Language Processing (NLP) system for detecting several types of spin in biomedical articles reporting randomized controlled trials (RCTs). We use a combination of rule-based and machine learning approaches to extract important information on trial design and to detect potential spin. The proposed spin detection system includes algorithms for text structure analysis, sentence classification, entity and relation extraction, semantic similarity assessment. Our algorithms achieved operational performance for the these tasks, F-measure ranging from 79,42 to 97.86% for different tasks. The most difficult task is extracting reported outcomes. Our tool is intended to be used as a semi-automated aid tool for assisting both authors and peer reviewers to detect potential spin. The tool incorporates a simple interface that allows to run the algorithms and visualize their output. It can also be used for manual annotation and correction of the errors in the outputs. The proposed tool is the first tool for spin detection. The tool and the annotated dataset are freely available.
{"title":"DeSpin: a prototype system for detecting spin in biomedical publications","authors":"A. Koroleva, S. Kamath, P. Bossuyt, P. Paroubek","doi":"10.18653/v1/2020.bionlp-1.5","DOIUrl":"https://doi.org/10.18653/v1/2020.bionlp-1.5","url":null,"abstract":"Improving the quality of medical research reporting is crucial to reduce avoidable waste in research and to improve the quality of health care. Despite various initiatives aiming at improving research reporting – guidelines, checklists, authoring aids, peer review procedures, etc. – overinterpretation of research results, also known as spin, is still a serious issue in research reporting. In this paper, we propose a Natural Language Processing (NLP) system for detecting several types of spin in biomedical articles reporting randomized controlled trials (RCTs). We use a combination of rule-based and machine learning approaches to extract important information on trial design and to detect potential spin. The proposed spin detection system includes algorithms for text structure analysis, sentence classification, entity and relation extraction, semantic similarity assessment. Our algorithms achieved operational performance for the these tasks, F-measure ranging from 79,42 to 97.86% for different tasks. The most difficult task is extracting reported outcomes. Our tool is intended to be used as a semi-automated aid tool for assisting both authors and peer reviewers to detect potential spin. The tool incorporates a simple interface that allows to run the algorithms and visualize their output. It can also be used for manual annotation and correction of the errors in the outputs. The proposed tool is the first tool for spin detection. The tool and the annotated dataset are freely available.","PeriodicalId":200974,"journal":{"name":"Workshop on Biomedical Natural Language Processing","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121985433","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-07-01DOI: 10.18653/v1/2020.bionlp-1.11
Yuxia Wang, Fei Liu, Karin M. Verspoor, Timothy Baldwin
In this paper, we apply pre-trained language models to the Semantic Textual Similarity (STS) task, with a specific focus on the clinical domain. In low-resource setting of clinical STS, these large models tend to be impractical and prone to overfitting. Building on BERT, we study the impact of a number of model design choices, namely different fine-tuning and pooling strategies. We observe that the impact of domain-specific fine-tuning on clinical STS is much less than that in the general domain, likely due to the concept richness of the domain. Based on this, we propose two data augmentation techniques. Experimental results on N2C2-STS 1 demonstrate substantial improvements, validating the utility of the proposed methods.
{"title":"Evaluating the Utility of Model Configurations and Data Augmentation on Clinical Semantic Textual Similarity","authors":"Yuxia Wang, Fei Liu, Karin M. Verspoor, Timothy Baldwin","doi":"10.18653/v1/2020.bionlp-1.11","DOIUrl":"https://doi.org/10.18653/v1/2020.bionlp-1.11","url":null,"abstract":"In this paper, we apply pre-trained language models to the Semantic Textual Similarity (STS) task, with a specific focus on the clinical domain. In low-resource setting of clinical STS, these large models tend to be impractical and prone to overfitting. Building on BERT, we study the impact of a number of model design choices, namely different fine-tuning and pooling strategies. We observe that the impact of domain-specific fine-tuning on clinical STS is much less than that in the general domain, likely due to the concept richness of the domain. Based on this, we propose two data augmentation techniques. Experimental results on N2C2-STS 1 demonstrate substantial improvements, validating the utility of the proposed methods.","PeriodicalId":200974,"journal":{"name":"Workshop on Biomedical Natural Language Processing","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126073143","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：