Opinion question answering is a challenging task for natural language processing. In this paper, we discuss a necessary component for an opinion question answering system: separating opinions from fact, at both the document and sentence level. We present a Bayesian classifier for discriminating between documents with a preponderance of opinions such as editorials from regular news stories, and describe three unsupervised, statistical techniques for the significantly harder task of detecting opinions at the sentence level. We also present a first model for classifying opinion sentences as positive or negative in terms of the main perspective being expressed in the opinion. Results from a large collection of news stories and a human evaluation of 400 sentences are reported, indicating that we achieve very high performance in document classification (upwards of 97% precision and recall), and respectable performance in detecting opinions and classifying them at the sentence level as positive, negative, or neutral (up to 91% accuracy).
{"title":"Towards Answering Opinion Questions: Separating Facts from Opinions and Identifying the Polarity of Opinion Sentences.","authors":"Hong Yu, Vasileios Hatzivassiloglou","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>Opinion question answering is a challenging task for natural language processing. In this paper, we discuss a necessary component for an opinion question answering system: separating opinions from fact, at both the document and sentence level. We present a Bayesian classifier for discriminating between documents with a preponderance of opinions such as editorials from regular news stories, and describe three unsupervised, statistical techniques for the significantly harder task of detecting opinions at the sentence level. We also present a first model for classifying opinion sentences as positive or negative in terms of the main perspective being expressed in the opinion. Results from a large collection of news stories and a human evaluation of 400 sentences are reported, indicating that we achieve very high performance in document classification (upwards of 97% precision and recall), and respectable performance in detecting opinions and classifying them at the sentence level as positive, negative, or neutral (up to 91% accuracy).</p>","PeriodicalId":74540,"journal":{"name":"Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing","volume":"2003 ","pages":"129-136"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12836483/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146095221","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-26DOI: 10.48550/arXiv.2212.13036
Yechun Tang, Xiaoxia Cheng, Weiming Lu
Complex knowledge base question answering can be achieved by converting questions into sequences of predefined actions. However, there is a significant semantic and structural gap between natural language and action sequences, which makes this conversion difficult. In this paper, we introduce an alignment-enhanced complex question answering framework, called ALCQA, which mitigates this gap through question-to-action alignment and question-to-question alignment. We train a question rewriting model to align the question and each action, and utilize a pretrained language model to implicitly align the question and KG artifacts. Moreover, considering that similar questions correspond to similar action sequences, we retrieve top-k similar question-answer pairs at the inference stage through question-to-question alignment and propose a novel reward-guided action sequence selection strategy to select from candidate action sequences. We conduct experiments on CQA and WQSP datasets, and the results show that our approach outperforms state-of-the-art methods and obtains a 9.88% improvements in the F1 metric on CQA dataset. Our source code is available at https://github.com/TTTTTTTTy/ALCQA.
{"title":"Improving Complex Knowledge Base Question Answering via Question-to-Action and Question-to-Question Alignment","authors":"Yechun Tang, Xiaoxia Cheng, Weiming Lu","doi":"10.48550/arXiv.2212.13036","DOIUrl":"https://doi.org/10.48550/arXiv.2212.13036","url":null,"abstract":"Complex knowledge base question answering can be achieved by converting questions into sequences of predefined actions. However, there is a significant semantic and structural gap between natural language and action sequences, which makes this conversion difficult. In this paper, we introduce an alignment-enhanced complex question answering framework, called ALCQA, which mitigates this gap through question-to-action alignment and question-to-question alignment. We train a question rewriting model to align the question and each action, and utilize a pretrained language model to implicitly align the question and KG artifacts. Moreover, considering that similar questions correspond to similar action sequences, we retrieve top-k similar question-answer pairs at the inference stage through question-to-question alignment and propose a novel reward-guided action sequence selection strategy to select from candidate action sequences. We conduct experiments on CQA and WQSP datasets, and the results show that our approach outperforms state-of-the-art methods and obtains a 9.88% improvements in the F1 metric on CQA dataset. Our source code is available at https://github.com/TTTTTTTTy/ALCQA.","PeriodicalId":74540,"journal":{"name":"Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing","volume":"52 1","pages":"137-147"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84959619","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 : 2022-12-26DOI: 10.48550/arXiv.2212.13005
Tianyi Tang, Junyi Li, Z. Chen, Yiwen Hu, Zhuohao Yu, Wen-Dao Dai, Zican Dong, Xiaoxue Cheng, Yuhao Wang, Wayne Xin Zhao, J. Nie, Ji-rong Wen
To facilitate research on text generation, this paper presents a comprehensive and unified library, TextBox 2.0, focusing on the use of pre-trained language models (PLMs). To be comprehensive, our library covers $13$ common text generation tasks and their corresponding $83$ datasets and further incorporates $45$ PLMs covering general, translation, Chinese, dialogue, controllable, distilled, prompting, and lightweight PLMs. We also implement $4$ efficient training strategies and provide $4$ generation objectives for pre-training new PLMs from scratch. To be unified, we design the interfaces to support the entire research pipeline (from data loading to training and evaluation), ensuring that each step can be fulfilled in a unified way. Despite the rich functionality, it is easy to use our library, either through the friendly Python API or command line. To validate the effectiveness of our library, we conduct extensive experiments and exemplify four types of research scenarios. The project is released at the link: https://github.com/RUCAIBox/TextBox.
{"title":"TextBox 2.0: A Text Generation Library with Pre-trained Language Models","authors":"Tianyi Tang, Junyi Li, Z. Chen, Yiwen Hu, Zhuohao Yu, Wen-Dao Dai, Zican Dong, Xiaoxue Cheng, Yuhao Wang, Wayne Xin Zhao, J. Nie, Ji-rong Wen","doi":"10.48550/arXiv.2212.13005","DOIUrl":"https://doi.org/10.48550/arXiv.2212.13005","url":null,"abstract":"To facilitate research on text generation, this paper presents a comprehensive and unified library, TextBox 2.0, focusing on the use of pre-trained language models (PLMs). To be comprehensive, our library covers $13$ common text generation tasks and their corresponding $83$ datasets and further incorporates $45$ PLMs covering general, translation, Chinese, dialogue, controllable, distilled, prompting, and lightweight PLMs. We also implement $4$ efficient training strategies and provide $4$ generation objectives for pre-training new PLMs from scratch. To be unified, we design the interfaces to support the entire research pipeline (from data loading to training and evaluation), ensuring that each step can be fulfilled in a unified way. Despite the rich functionality, it is easy to use our library, either through the friendly Python API or command line. To validate the effectiveness of our library, we conduct extensive experiments and exemplify four types of research scenarios. The project is released at the link: https://github.com/RUCAIBox/TextBox.","PeriodicalId":74540,"journal":{"name":"Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing","volume":"452 1","pages":"435-444"},"PeriodicalIF":0.0,"publicationDate":"2022-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76493273","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 : 2022-12-24DOI: 10.48550/arXiv.2212.12652
Borui Wang, Chengcheng Feng, Arjun Nair, Madelyn Mao, Jai Desai, Asli Celikyilmaz, Haoran Li, Yashar Mehdad, Dragomir R. Radev
Abstractive dialogue summarization has long been viewed as an important standalone task in natural language processing, but no previous work has explored the possibility of whether abstractive dialogue summarization can also be used as a means to boost an NLP system’s performance on other important dialogue comprehension tasks. In this paper, we propose a novel type of dialogue summarization task - STRUctured DiaLoguE Summarization (STRUDEL) - that can help pre-trained language models to better understand dialogues and improve their performance on important dialogue comprehension tasks. In contrast to the holistic approach taken by the traditional free-form abstractive summarization task for dialogues, STRUDEL aims to decompose and imitate the hierarchical, systematic and structured mental process that we human beings usually go through when understanding and analyzing dialogues, and thus has the advantage of being more focused, specific and instructive for dialogue comprehension models to learn from. We further introduce a new STRUDEL dialogue comprehension modeling framework that integrates STRUDEL into a dialogue reasoning module over transformer encoder language models to improve their dialogue comprehension ability. In our empirical experiments on two important downstream dialogue comprehension tasks - dialogue question answering and dialogue response prediction - we demonstrate that our STRUDEL dialogue comprehension models can significantly improve the dialogue comprehension performance of transformer encoder language models.
{"title":"STRUDEL: Structured Dialogue Summarization for Dialogue Comprehension","authors":"Borui Wang, Chengcheng Feng, Arjun Nair, Madelyn Mao, Jai Desai, Asli Celikyilmaz, Haoran Li, Yashar Mehdad, Dragomir R. Radev","doi":"10.48550/arXiv.2212.12652","DOIUrl":"https://doi.org/10.48550/arXiv.2212.12652","url":null,"abstract":"Abstractive dialogue summarization has long been viewed as an important standalone task in natural language processing, but no previous work has explored the possibility of whether abstractive dialogue summarization can also be used as a means to boost an NLP system’s performance on other important dialogue comprehension tasks. In this paper, we propose a novel type of dialogue summarization task - STRUctured DiaLoguE Summarization (STRUDEL) - that can help pre-trained language models to better understand dialogues and improve their performance on important dialogue comprehension tasks. In contrast to the holistic approach taken by the traditional free-form abstractive summarization task for dialogues, STRUDEL aims to decompose and imitate the hierarchical, systematic and structured mental process that we human beings usually go through when understanding and analyzing dialogues, and thus has the advantage of being more focused, specific and instructive for dialogue comprehension models to learn from. We further introduce a new STRUDEL dialogue comprehension modeling framework that integrates STRUDEL into a dialogue reasoning module over transformer encoder language models to improve their dialogue comprehension ability. In our empirical experiments on two important downstream dialogue comprehension tasks - dialogue question answering and dialogue response prediction - we demonstrate that our STRUDEL dialogue comprehension models can significantly improve the dialogue comprehension performance of transformer encoder language models.","PeriodicalId":74540,"journal":{"name":"Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing","volume":"160 3 1","pages":"4949-4958"},"PeriodicalIF":0.0,"publicationDate":"2022-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83262235","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 : 2022-12-21DOI: 10.48550/arXiv.2212.10879
Ningyu Xu, Tao Gui, Ruotian Ma, Qi Zhang, Jingting Ye, Menghan Zhang, Xuanjing Huang
Multilingual BERT (mBERT) has demonstrated considerable cross-lingual syntactic ability, whereby it enables effective zero-shot cross-lingual transfer of syntactic knowledge. The transfer is more successful between some languages, but it is not well understood what leads to this variation and whether it fairly reflects difference between languages. In this work, we investigate the distributions of grammatical relations induced from mBERT in the context of 24 typologically different languages. We demonstrate that the distance between the distributions of different languages is highly consistent with the syntactic difference in terms of linguistic formalisms. Such difference learnt via self-supervision plays a crucial role in the zero-shot transfer performance and can be predicted by variation in morphosyntactic properties between languages. These results suggest that mBERT properly encodes languages in a way consistent with linguistic diversity and provide insights into the mechanism of cross-lingual transfer.
{"title":"Cross-Linguistic Syntactic Difference in Multilingual BERT: How Good is It and How Does It Affect Transfer?","authors":"Ningyu Xu, Tao Gui, Ruotian Ma, Qi Zhang, Jingting Ye, Menghan Zhang, Xuanjing Huang","doi":"10.48550/arXiv.2212.10879","DOIUrl":"https://doi.org/10.48550/arXiv.2212.10879","url":null,"abstract":"Multilingual BERT (mBERT) has demonstrated considerable cross-lingual syntactic ability, whereby it enables effective zero-shot cross-lingual transfer of syntactic knowledge. The transfer is more successful between some languages, but it is not well understood what leads to this variation and whether it fairly reflects difference between languages. In this work, we investigate the distributions of grammatical relations induced from mBERT in the context of 24 typologically different languages. We demonstrate that the distance between the distributions of different languages is highly consistent with the syntactic difference in terms of linguistic formalisms. Such difference learnt via self-supervision plays a crucial role in the zero-shot transfer performance and can be predicted by variation in morphosyntactic properties between languages. These results suggest that mBERT properly encodes languages in a way consistent with linguistic diversity and provide insights into the mechanism of cross-lingual transfer.","PeriodicalId":74540,"journal":{"name":"Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing","volume":"71 1","pages":"8073-8092"},"PeriodicalIF":0.0,"publicationDate":"2022-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74047386","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 : 2022-12-21DOI: 10.48550/arXiv.2212.10843
Dongmin Hyun, Xiting Wang, Chanyoung Park, Xing Xie, Hwanjo Yu
Sentence summarization shortens given texts while maintaining core contents of the texts. Unsupervised approaches have been studied to summarize texts without human-written summaries. However, recent unsupervised models are extractive, which remove words from texts and thus they are less flexible than abstractive summarization. In this work, we devise an abstractive model based on reinforcement learning without ground-truth summaries. We formulate the unsupervised summarization based on the Markov decision process with rewards representing the summary quality. To further enhance the summary quality, we develop a multi-summary learning mechanism that generates multiple summaries with varying lengths for a given text, while making the summaries mutually enhance each other. Experimental results show that the proposed model substantially outperforms both abstractive and extractive models, yet frequently generating new words not contained in input texts.
{"title":"Generating Multiple-Length Summaries via Reinforcement Learning for Unsupervised Sentence Summarization","authors":"Dongmin Hyun, Xiting Wang, Chanyoung Park, Xing Xie, Hwanjo Yu","doi":"10.48550/arXiv.2212.10843","DOIUrl":"https://doi.org/10.48550/arXiv.2212.10843","url":null,"abstract":"Sentence summarization shortens given texts while maintaining core contents of the texts. Unsupervised approaches have been studied to summarize texts without human-written summaries. However, recent unsupervised models are extractive, which remove words from texts and thus they are less flexible than abstractive summarization. In this work, we devise an abstractive model based on reinforcement learning without ground-truth summaries. We formulate the unsupervised summarization based on the Markov decision process with rewards representing the summary quality. To further enhance the summary quality, we develop a multi-summary learning mechanism that generates multiple summaries with varying lengths for a given text, while making the summaries mutually enhance each other. Experimental results show that the proposed model substantially outperforms both abstractive and extractive models, yet frequently generating new words not contained in input texts.","PeriodicalId":74540,"journal":{"name":"Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing","volume":"91 1","pages":"2939-2951"},"PeriodicalIF":0.0,"publicationDate":"2022-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90256373","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 : 2022-12-16DOI: 10.48550/arXiv.2212.08216
Gabrielle Gauthier Melançon, Orlando Marquez Ayala, Lindsay D. Brin, Chris Tyler, Frederic Branchaud-Charron, Joseph Marinier, Karine Grande, Dieu-Thu Le
We present Azimuth, an open-source and easy-to-use tool to perform error analysis for text classification. Compared to other stages of the ML development cycle, such as model training and hyper-parameter tuning, the process and tooling for the error analysis stage are less mature. However, this stage is critical for the development of reliable and trustworthy AI systems. To make error analysis more systematic, we propose an approach comprising dataset analysis and model quality assessment, which Azimuth facilitates. We aim to help AI practitioners discover and address areas where the model does not generalize by leveraging and integrating a range of ML techniques, such as saliency maps, similarity, uncertainty, and behavioral analyses, all in one tool. Our code and documentation are available at github.com/servicenow/azimuth.
{"title":"Azimuth: Systematic Error Analysis for Text Classification","authors":"Gabrielle Gauthier Melançon, Orlando Marquez Ayala, Lindsay D. Brin, Chris Tyler, Frederic Branchaud-Charron, Joseph Marinier, Karine Grande, Dieu-Thu Le","doi":"10.48550/arXiv.2212.08216","DOIUrl":"https://doi.org/10.48550/arXiv.2212.08216","url":null,"abstract":"We present Azimuth, an open-source and easy-to-use tool to perform error analysis for text classification. Compared to other stages of the ML development cycle, such as model training and hyper-parameter tuning, the process and tooling for the error analysis stage are less mature. However, this stage is critical for the development of reliable and trustworthy AI systems. To make error analysis more systematic, we propose an approach comprising dataset analysis and model quality assessment, which Azimuth facilitates. We aim to help AI practitioners discover and address areas where the model does not generalize by leveraging and integrating a range of ML techniques, such as saliency maps, similarity, uncertainty, and behavioral analyses, all in one tool. Our code and documentation are available at github.com/servicenow/azimuth.","PeriodicalId":74540,"journal":{"name":"Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing","volume":"71 1","pages":"298-310"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84669864","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 : 2022-12-16DOI: 10.48550/arXiv.2212.13898
Jai Gupta, Yi Tay, C. Kamath, Vinh Q. Tran, Donald Metzler, S. Bavadekar, Mimi Sun, E. Gabrilovich
With the devastating outbreak of COVID-19, vaccines are one of the crucial lines of defense against mass infection in this global pandemic. Given the protection they provide, vaccines are becoming mandatory in certain social and professional settings. This paper presents a classification model for detecting COVID-19 vaccination related search queries, a machine learning model that is used to generate search insights for COVID-19 vaccinations. The proposed method combines and leverages advancements from modern state-of-the-art (SOTA) natural language understanding (NLU) techniques such as pretrained Transformers with traditional dense features. We propose a novel approach of considering dense features as memory tokens that the model can attend to. We show that this new modeling approach enables a significant improvement to the Vaccine Search Insights (VSI) task, improving a strong well-established gradient-boosting baseline by relative +15% improvement in F1 score and +14% in precision.
{"title":"Dense Feature Memory Augmented Transformers for COVID-19 Vaccination Search Classification","authors":"Jai Gupta, Yi Tay, C. Kamath, Vinh Q. Tran, Donald Metzler, S. Bavadekar, Mimi Sun, E. Gabrilovich","doi":"10.48550/arXiv.2212.13898","DOIUrl":"https://doi.org/10.48550/arXiv.2212.13898","url":null,"abstract":"With the devastating outbreak of COVID-19, vaccines are one of the crucial lines of defense against mass infection in this global pandemic. Given the protection they provide, vaccines are becoming mandatory in certain social and professional settings. This paper presents a classification model for detecting COVID-19 vaccination related search queries, a machine learning model that is used to generate search insights for COVID-19 vaccinations. The proposed method combines and leverages advancements from modern state-of-the-art (SOTA) natural language understanding (NLU) techniques such as pretrained Transformers with traditional dense features. We propose a novel approach of considering dense features as memory tokens that the model can attend to. We show that this new modeling approach enables a significant improvement to the Vaccine Search Insights (VSI) task, improving a strong well-established gradient-boosting baseline by relative +15% improvement in F1 score and +14% in precision.","PeriodicalId":74540,"journal":{"name":"Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing","volume":"43 1","pages":"521-530"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84126133","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 : 2022-12-16DOI: 10.48550/arXiv.2212.08322
Kai Xiong, Xiao Ding, Zhongyang Li, L. Du, Bing Qin, Yi Zheng, Baoxing Huai
Causal chain reasoning (CCR) is an essential ability for many decision-making AI systems, which requires the model to build reliable causal chains by connecting causal pairs. However, CCR suffers from two main transitive problems: threshold effect and scene drift. In other words, the causal pairs to be spliced may have a conflicting threshold boundary or scenario.To address these issues, we propose a novel Reliable Causal chain reasoning framework (ReCo), which introduces exogenous variables to represent the threshold and scene factors of each causal pair within the causal chain, and estimates the threshold and scene contradictions across exogenous variables via structural causal recurrent neural networks (SRNN). Experiments show that ReCo outperforms a series of strong baselines on both Chinese and English CCR datasets. Moreover, by injecting reliable causal chain knowledge distilled by ReCo, BERT can achieve better performances on four downstream causal-related tasks than BERT models enhanced by other kinds of knowledge.
{"title":"ReCo: Reliable Causal Chain Reasoning via Structural Causal Recurrent Neural Networks","authors":"Kai Xiong, Xiao Ding, Zhongyang Li, L. Du, Bing Qin, Yi Zheng, Baoxing Huai","doi":"10.48550/arXiv.2212.08322","DOIUrl":"https://doi.org/10.48550/arXiv.2212.08322","url":null,"abstract":"Causal chain reasoning (CCR) is an essential ability for many decision-making AI systems, which requires the model to build reliable causal chains by connecting causal pairs. However, CCR suffers from two main transitive problems: threshold effect and scene drift. In other words, the causal pairs to be spliced may have a conflicting threshold boundary or scenario.To address these issues, we propose a novel Reliable Causal chain reasoning framework (ReCo), which introduces exogenous variables to represent the threshold and scene factors of each causal pair within the causal chain, and estimates the threshold and scene contradictions across exogenous variables via structural causal recurrent neural networks (SRNN). Experiments show that ReCo outperforms a series of strong baselines on both Chinese and English CCR datasets. Moreover, by injecting reliable causal chain knowledge distilled by ReCo, BERT can achieve better performances on four downstream causal-related tasks than BERT models enhanced by other kinds of knowledge.","PeriodicalId":74540,"journal":{"name":"Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing","volume":"30 1","pages":"6426-6438"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86496573","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 : 2022-12-15DOI: 10.48550/arXiv.2212.08120
Denis Emelin, Daniele Bonadiman, Sawsan Alqahtani, Yi Zhang, Saab Mansour
Pre-trained language models (PLM) have advanced the state-of-the-art across NLP applications, but lack domain-specific knowledge that does not naturally occur in pre-training data. Previous studies augmented PLMs with symbolic knowledge for different downstream NLP tasks. However, knowledge bases (KBs) utilized in these studies are usually large-scale and static, in contrast to small, domain-specific, and modifiable knowledge bases that are prominent in real-world task-oriented dialogue (TOD) systems. In this paper, we showcase the advantages of injecting domain-specific knowledge prior to fine-tuning on TOD tasks. To this end, we utilize light-weight adapters that can be easily integrated with PLMs and serve as a repository for facts learned from different KBs. To measure the efficacy of proposed knowledge injection methods, we introduce Knowledge Probing using Response Selection (KPRS) – a probe designed specifically for TOD models. Experiments on KPRS and the response generation task show improvements of knowledge injection with adapters over strong baselines.
{"title":"Injecting Domain Knowledge in Language Models for Task-oriented Dialogue Systems","authors":"Denis Emelin, Daniele Bonadiman, Sawsan Alqahtani, Yi Zhang, Saab Mansour","doi":"10.48550/arXiv.2212.08120","DOIUrl":"https://doi.org/10.48550/arXiv.2212.08120","url":null,"abstract":"Pre-trained language models (PLM) have advanced the state-of-the-art across NLP applications, but lack domain-specific knowledge that does not naturally occur in pre-training data. Previous studies augmented PLMs with symbolic knowledge for different downstream NLP tasks. However, knowledge bases (KBs) utilized in these studies are usually large-scale and static, in contrast to small, domain-specific, and modifiable knowledge bases that are prominent in real-world task-oriented dialogue (TOD) systems. In this paper, we showcase the advantages of injecting domain-specific knowledge prior to fine-tuning on TOD tasks. To this end, we utilize light-weight adapters that can be easily integrated with PLMs and serve as a repository for facts learned from different KBs. To measure the efficacy of proposed knowledge injection methods, we introduce Knowledge Probing using Response Selection (KPRS) – a probe designed specifically for TOD models. Experiments on KPRS and the response generation task show improvements of knowledge injection with adapters over strong baselines.","PeriodicalId":74540,"journal":{"name":"Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing","volume":"56 1","pages":"11962-11974"},"PeriodicalIF":0.0,"publicationDate":"2022-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79138347","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号