Events and Stories in the News

Computational Representation of News Events and Storylines

Events and Stories in the News is a workshop series and multidisciplinary research community investigating technologies for representing and accessing events and storylines from news as data structures. This approach offers an alternative to capturing events and stories using natural language artifacts, such as text articles, reports and videos. Understanding events and stories as data presents significant challenges that touch on semantics, natural language processing, knowledge engineering, artificial intelligence, human cognition and even philosophy.

Today’s digital media ecosystem generates massive streams of news, largely in the form of individual documents within which news events and narrative structures are communicated using natural language text. The increasing quantity of text documents produced by the ecosystem presents challenges to those seeking to understand and contextualize events and narratives over long periods of time, leading to demands for new multidimensional, multimodal and distributed representations of news events and of the structures involved in “stories”.

Stories are essentially sequences of events with an internal coherence: being about the protagonists, observing a timeline, tracking the states and changes in the world as related to the protagonists, and (usually, for good stories) having some kind of point or moral. Understanding and representing stories require addressing additional areas, including: i.) state detection and state change tracking across time; ii.) storyline analysis and moral/point determination; iii.) multi-source and multimedia integration; iv.) perspective identification and perspective-based analysis and summarization of stories. Some of these focus areas are already being addressed in computational terms, others are still very much purely in the realm of linguistics, professional writing, and journalism.

Furthermore, NLP technologies still struggle with, among others, event detection, event identity and coreference, and causal, topical, temporal, and spatial relations between events. People, on the other hand, easily decide on salient events, such as changes in the world, make reference to these events and their participants through linguistic expressions, and construct storylines using only the pertinent aspects. Such narrative structures are at the heart of information sharing, as is exemplified by the structure of news articles. But it remains extraordinarily difficult to detect them automatically, let alone to automatically construct stories from such event representations.

This website provides information about the current year’s workshop. It also provides links to websites of previous workshops in the series and to websites of related gatherings, including access to proceedings where possible. It also lists related resources, such as relevant journal articles, books, organizations, etc. We invite interested researchers to participate in this community by submitting to, or attending, the workshop.

Annotation Task
The 2018 workshop will include an exercise focused on annotating events and storylines. Preparation prior to the workshop is recommended.

A description and instructions for the annotation exercise can be found here.

Workshop Program
The program for the 2018 workshop has been finalized.

The full schedule is available here.

CFP for the 2018 Workshop
The 2018 workship will be co-located with COLING 2018, in Santa Fe, New Mexico, USA, on Monday August 20th.

The CFP announcement is available here.