key: cord-0039589-7ryitk6y authors: Campos, Ricardo; Jorge, Alí­pio; Jatowt, Adam; Bhatia, Sumit title: The 3[Formula: see text] International Workshop on Narrative Extraction from Texts: Text2Story 2020 date: 2020-03-24 journal: Advances in Information Retrieval DOI: 10.1007/978-3-030-45442-5_86 sha: ae137e7eccec64229cb992ce745f2e582d65f9d4 doc_id: 39589 cord_uid: 7ryitk6y The Third International Workshop on Narrative Extraction from Texts (Text2Story’20) [text2story20.inesctec.pt] held in conjunction with the 42[Formula: see text] European Conference on Information Retrieval (ECIR 2020) gives researchers of IR, NLP and other fields, the opportunity to share their recent advances in extraction and formal representation of narratives. This workshop also presents a forum to consolidate the multi-disciplinary efforts and foster discussions around the narrative extraction task, a hot topic in recent years. Searching for relevant information is a permanent need for those who want to stay informed about a given event, news or story. While having access to information is now easier than ever with the proliferation of devices and different means of accessing data, keeping up-to-date with all the developments and various aspects of the topic being followed is a difficult task. In many situations, it is hard for readers to connect the dots of a given story [21] . This is due not only to the widespread presence of the media outlets in the digital space [14] , but also to the increasing participation of citizens, who produce and promote an unprecedented number of comments and discussions on social media (some of them lasting over days, weeks or months). Automatic narrative extraction from texts offers a compelling approach to this problem by automatically identifying the sub-set of interconnected raw documents, extracting the critical narrative/story elements, and representing them in a more adequate manner that conveys the key points of the story in an easy to understand format to the readers. This could be done through text summarization [13] , timelines [15, 19] , word clouds [3, 20] , visual textual analytics [7, 18] or in an intermediate structured formalism (e.g., wikilike page structures [1] ) that can feed further steps (e.g., gamification [6, 16] or story generation [12] (such as automatically generating finance [5, 11] and sport reports [22] )). Although information extraction and natural language processing have made significant progress towards automatic interpretation of texts, the problem of fully identifying and relating the different elements of a narrative in a document (set) still presents significant unsolved challenges [17] . The purpose of the Text2Story workshop series is to shorten the distance between IR and people working on automatic narrative extraction and construction from texts, a vibrant line of research that has been conducted over the last few years by many research groups. The Text2Story workshop series had its first edition at ECIR'18 [8] , followed by a second edition on ECIR'19 [9] . In these two first editions, we had an approximate number of 70 participants, 16 research papers presented, plus demo and poster sessions, and vibrant talks from our four invited keynotes: Udo Kruschwitz (University of Essex), Eric Gaussier (University Grenoble Alps), Iryna Gurevych (Technische Universität Darmstadt), and Miguel Martinez-Alvarez (Signal AI). In addition to this, we also edited the Text2Story Special Issue on IPM Journal [10] which had more than 30 submissions and 8 papers accepted, demonstrating the growing activity of this specific research area. The organizers of the workshop have also been actively involved in this research area with the proposal and the contribution of new methods and solutions, most notably the YAKE! keyword extraction algorithm [2] [3] [4] -best short paper of ECIR'18 -and the Tell me Stories temporal summarization tool [15] -best demo presentation at ECIR'19. In the third edition of this workshop series, we aim to raise awareness to the problem of creating text-to-narrative-structures and its related tasks. We focus on researchers and practitioners working on identifying, extracting and producing narrative stories, but also on people from industry, particularly journalists and stakeholders working in traditional and social media. The call for papers aimed to cover original research at the intersection of IR and NLP on all aspects of storyline identification and generation from texts including but not limited to narrative and content generation, formal representation, and visualization of narratives. The topics of the workshop are in line with the previous editions of the Text2Story workshop series. In particular, we featured the following topics: How it happened: discovering and archiving the evolution of a story using social signals A text feature based automatic keyword extraction method for single documents YAKE! collection-independent automatic keyword extractor YAKE! keyword extraction from single documents using multiple local features The first financial narrative processing workshop (FNP 2018) Narrative gamification as a method of increasing sales performance: a field experimental study MultiConVis: a visual text analytics system for exploring a collection of online conversations First international workshop on narrative extraction from texts Second international workshop on narrative extraction from texts Special issue on narrative extraction from texts (Text2Story): preface. Inf. Process. Manage Fad or future? Automated analysis of financial text and its implications for corporate reporting Story generation with crowdsourced plot graphs TIARA: interactive, topic-based visual text summarization and analysis First international workshop on recent trends in news information retrieval Interactive system for automatically generating temporal narratives Stories or scenarios: implementing narratives in gamified language teaching Computational narrative intelligence: a human-centered goal for artificial intelligence Visual analysis and knowledge discovery for text Timeline summarization from relevant headlines Tag clouds and the case for vernacular visualization Storylines for structuring massive streams of news Towards constructing sports news from live text commentary Acknowledgements. The first two authors of this paper are financed by the ERDF -European Regional Development Fund through the North Portugal Regional Operational Programme (NORTE 2020), under the PORTUGAL 2020 and by National Funds through the Portuguese funding agency, FCT -Fundação para a Ciência e a Tecnologia within project PTDC/CCI-COM/31857/2017 (NORTE-01-0145-FEDER-03185).