Pub Date : 2021-11-30DOI: 10.1017/9781108854221.011
Heng Ji, Clare R. Voss
Event Extraction aims to find who did what to whom, when and where from unstructured data. Over the past decade, research in event extraction has made advances in three waves. The first wave relied on supervised machine learning models trained from a large amount of manually annotated data and manually crafted features. The second wave eliminated this method of feature engineering by introducing deep neural networks with distributional semantic embedding features, but still required large annotated datasets. This chapter provides an overview of a third wave with a share-and-transfer framework, that further enhances the portability of event extraction by transferring knowledge from a high-resource setting to another low-resource setting, reducing the need there for annotated data. We describe three low-resource settings: a new domain, a new language, or a new data modality. The first share step of our approach is to construct a common structured semantic representation space into which these complex structures can be encoded. Then, in the transfer step of the approach, we can train event extractors over these representations in high-resource settings and apply the learned extractors to target data in the low-resource setting. We conclude the a Supported by ARL NS-CTA No. W911NF-09-2-0053, DARPA KAIROS Program # FA8750-19-2-1004, U.S. DARPA LORELEI Program # HR0011-15-C-0115, U.S. DARPA AIDA Program # FA8750-18-2-0014, Air Force No. FA8650-17-C-7715, and the Office of the Director of National Intelligence (ODNI), Intelligence Advanced Research Projects Activity (IARPA), via contract # FA8650-17-C-9116. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied, of DARPA, ODNI, IARPA, or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for governmental purposes notwithstanding any copyright annotation therein. 2Low-resource Event Extraction via Share-and-Transfer and Remaining Challenges chapter with a summary of the current status of this new framework and point to remaining challenges and future research directions to address them.
{"title":"Low-Resource Event Extraction via Share-and-Transfer and Remaining Challenges","authors":"Heng Ji, Clare R. Voss","doi":"10.1017/9781108854221.011","DOIUrl":"https://doi.org/10.1017/9781108854221.011","url":null,"abstract":"Event Extraction aims to find who did what to whom, when and where from unstructured data. Over the past decade, research in event extraction has made advances in three waves. The first wave relied on supervised machine learning models trained from a large amount of manually annotated data and manually crafted features. The second wave eliminated this method of feature engineering by introducing deep neural networks with distributional semantic embedding features, but still required large annotated datasets. This chapter provides an overview of a third wave with a share-and-transfer framework, that further enhances the portability of event extraction by transferring knowledge from a high-resource setting to another low-resource setting, reducing the need there for annotated data. We describe three low-resource settings: a new domain, a new language, or a new data modality. The first share step of our approach is to construct a common structured semantic representation space into which these complex structures can be encoded. Then, in the transfer step of the approach, we can train event extractors over these representations in high-resource settings and apply the learned extractors to target data in the low-resource setting. We conclude the a Supported by ARL NS-CTA No. W911NF-09-2-0053, DARPA KAIROS Program # FA8750-19-2-1004, U.S. DARPA LORELEI Program # HR0011-15-C-0115, U.S. DARPA AIDA Program # FA8750-18-2-0014, Air Force No. FA8650-17-C-7715, and the Office of the Director of National Intelligence (ODNI), Intelligence Advanced Research Projects Activity (IARPA), via contract # FA8650-17-C-9116. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied, of DARPA, ODNI, IARPA, or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for governmental purposes notwithstanding any copyright annotation therein. 2Low-resource Event Extraction via Share-and-Transfer and Remaining Challenges chapter with a summary of the current status of this new framework and point to remaining challenges and future research directions to address them.","PeriodicalId":170332,"journal":{"name":"Computational Analysis of Storylines","volume":"65 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120874815","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 : 2021-11-30DOI: 10.1017/9781108854221.012
{"title":"Reading Certainty across Sources","authors":"","doi":"10.1017/9781108854221.012","DOIUrl":"https://doi.org/10.1017/9781108854221.012","url":null,"abstract":"","PeriodicalId":170332,"journal":{"name":"Computational Analysis of Storylines","volume":"229 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124536014","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 : 2021-11-30DOI: 10.1017/9781108854221.010
{"title":"The Richer Event Description Corpus for Event–Event Relations","authors":"","doi":"10.1017/9781108854221.010","DOIUrl":"https://doi.org/10.1017/9781108854221.010","url":null,"abstract":"","PeriodicalId":170332,"journal":{"name":"Computational Analysis of Storylines","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115610187","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 : 2021-11-30DOI: 10.1017/9781108854221.015
{"title":"Semantic Storytelling: From Experiments and Prototypes to a Technical Solution","authors":"","doi":"10.1017/9781108854221.015","DOIUrl":"https://doi.org/10.1017/9781108854221.015","url":null,"abstract":"","PeriodicalId":170332,"journal":{"name":"Computational Analysis of Storylines","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116126988","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 : 2021-11-30DOI: 10.1017/9781108854221.006
Mark A. Finlayson, Andres Cremisini, M. Ocal
. Understanding the timeline of a story is a necessary first step for extracting storylines. This is difficult, because timelines are not explicitly given in documents, and parts of a story may be found across multiple documents, either repeated or in fragments. We outline prior work and the state of the art in both timeline extraction and alignment of timelines across documents. With regard to timeline extraction, there has been significant work over the past 40 years on representing temporal information in text, but most of it has focused on temporal graphs and not timelines. In the past 15 years researchers have begun to consider the problem of extracting timelines from these graphs, but the approaches have been incomplete and inexact. We review these approaches and describe recent work of our own that solves timeline extraction exactly. With regard to timeline alignment, most efforts have been focused only on the specific task of cross-document event coreference (CDEC). Current approaches to CDEC fall into two camps: event–only clustering and joint event–entity clustering, with joint clustering using neural methods achieving state-of-the-art performance. All CDEC approaches rely on document clustering to generate a tractable search space. We note both shortcomings and advantages of these various approaches and, importantly, we describe how CDEC falls short of full timeline alignment extraction. We outline next steps to advance the field toward full timeline alignment across documents that can serve as a foundation for extraction of higher-level, more abstract storylines.
{"title":"Extracting and Aligning Timelines","authors":"Mark A. Finlayson, Andres Cremisini, M. Ocal","doi":"10.1017/9781108854221.006","DOIUrl":"https://doi.org/10.1017/9781108854221.006","url":null,"abstract":". Understanding the timeline of a story is a necessary first step for extracting storylines. This is difficult, because timelines are not explicitly given in documents, and parts of a story may be found across multiple documents, either repeated or in fragments. We outline prior work and the state of the art in both timeline extraction and alignment of timelines across documents. With regard to timeline extraction, there has been significant work over the past 40 years on representing temporal information in text, but most of it has focused on temporal graphs and not timelines. In the past 15 years researchers have begun to consider the problem of extracting timelines from these graphs, but the approaches have been incomplete and inexact. We review these approaches and describe recent work of our own that solves timeline extraction exactly. With regard to timeline alignment, most efforts have been focused only on the specific task of cross-document event coreference (CDEC). Current approaches to CDEC fall into two camps: event–only clustering and joint event–entity clustering, with joint clustering using neural methods achieving state-of-the-art performance. All CDEC approaches rely on document clustering to generate a tractable search space. We note both shortcomings and advantages of these various approaches and, importantly, we describe how CDEC falls short of full timeline alignment extraction. We outline next steps to advance the field toward full timeline alignment across documents that can serve as a foundation for extraction of higher-level, more abstract storylines.","PeriodicalId":170332,"journal":{"name":"Computational Analysis of Storylines","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131762309","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 : 2021-11-30DOI: 10.1017/9781108854221.008
Piek Vossen, Tommaso Caselli, R. Segers
. Stories are a pervasive phenomenon of human life. They also represent a cognitive tool to understand and make sense of the world and of its happenings. In this contribution we describe a narratology-based framework for modeling stories as a combination of different data structures and to automatically extract them from news articles. We introduce a distinction among three data structures (timelines, causelines, and storylines) that capture different narratological dimensions, respectively chronological ordering, causal connections, and plot structure. We developed the Circumstantial Event Ontology (CEO) for modeling (implicit) circumstantial relations as well as explicit causal relations and create two benchmark corpora: ECB + / CEO, for causelines, and the Event Storyline Corpus (ESC), for storylines. To test our framework and the difficulty in automatically extract causelines and storylines, we develop a series of reasonable baseline systems.
{"title":"A Narratology-Based Framework for Storyline Extraction","authors":"Piek Vossen, Tommaso Caselli, R. Segers","doi":"10.1017/9781108854221.008","DOIUrl":"https://doi.org/10.1017/9781108854221.008","url":null,"abstract":". Stories are a pervasive phenomenon of human life. They also represent a cognitive tool to understand and make sense of the world and of its happenings. In this contribution we describe a narratology-based framework for modeling stories as a combination of different data structures and to automatically extract them from news articles. We introduce a distinction among three data structures (timelines, causelines, and storylines) that capture different narratological dimensions, respectively chronological ordering, causal connections, and plot structure. We developed the Circumstantial Event Ontology (CEO) for modeling (implicit) circumstantial relations as well as explicit causal relations and create two benchmark corpora: ECB + / CEO, for causelines, and the Event Storyline Corpus (ESC), for storylines. To test our framework and the difficulty in automatically extract causelines and storylines, we develop a series of reasonable baseline systems.","PeriodicalId":170332,"journal":{"name":"Computational Analysis of Storylines","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126434714","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 : 2021-11-30DOI: 10.1017/9781108854221.013
{"title":"Narrative Homogeneity and Heterogeneity in Document Categories","authors":"","doi":"10.1017/9781108854221.013","DOIUrl":"https://doi.org/10.1017/9781108854221.013","url":null,"abstract":"","PeriodicalId":170332,"journal":{"name":"Computational Analysis of Storylines","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124543836","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 : 2021-11-30DOI: 10.1017/9781108854221.003
J. Pustejovsky
. This chapter briefly reviews the research conducted on the representation of events, from the perspectives of natural language processing, artificial intelligence (AI), and linguistics. AI approaches to modeling change have traditionally focused on situations and state descriptions. Linguistic approaches start with the description of the propositional content of sentences (or natural language expressions generally). As a result, the focus in the two fields has been on different problems. Namely, linguistic theories try to maintain compositionality in the expressions associated with linguistic units, or what is known as semantic compositionality . In AI and in the planning community in particular the focus has been on maintaining compositionality in the way plans are constructed, as well as the correctness of the algorithm that searches and traverses the state space. This can be called plan compositionality . I argue that these approaches have common elements that can be drawn on to view event semantics from a unifying perspective, where we can distinguish between the surface events denoted by verbal predicates and what I refer to as the latent event structure of a sentence. Latent events within a text refer to the finer-grained subeventual representations of events denoted by verbs or nominal expressions, as well as to hidden events connoted by nouns. By clearly distinguishing between surface and latent event structures of sentences and texts, we move closer to a general computational theory of event structure, one permitting a common vocabulary for events and the relations between them, while enabling reasoning at multiple levels of interpretation.
{"title":"The Role of Event-Based Representations and Reasoning in Language","authors":"J. Pustejovsky","doi":"10.1017/9781108854221.003","DOIUrl":"https://doi.org/10.1017/9781108854221.003","url":null,"abstract":". This chapter briefly reviews the research conducted on the representation of events, from the perspectives of natural language processing, artificial intelligence (AI), and linguistics. AI approaches to modeling change have traditionally focused on situations and state descriptions. Linguistic approaches start with the description of the propositional content of sentences (or natural language expressions generally). As a result, the focus in the two fields has been on different problems. Namely, linguistic theories try to maintain compositionality in the expressions associated with linguistic units, or what is known as semantic compositionality . In AI and in the planning community in particular the focus has been on maintaining compositionality in the way plans are constructed, as well as the correctness of the algorithm that searches and traverses the state space. This can be called plan compositionality . I argue that these approaches have common elements that can be drawn on to view event semantics from a unifying perspective, where we can distinguish between the surface events denoted by verbal predicates and what I refer to as the latent event structure of a sentence. Latent events within a text refer to the finer-grained subeventual representations of events denoted by verbs or nominal expressions, as well as to hidden events connoted by nouns. By clearly distinguishing between surface and latent event structures of sentences and texts, we move closer to a general computational theory of event structure, one permitting a common vocabulary for events and the relations between them, while enabling reasoning at multiple levels of interpretation.","PeriodicalId":170332,"journal":{"name":"Computational Analysis of Storylines","volume":"121 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116290436","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：