Finding story chains in newswire articles

Abstract

Massive amounts of information about news events are published on the Internet every day in online newspapers, blogs, and social network messages. While search engines like Google help retrieve information using keywords, the large volumes of unstructured search results returned by search engines make it hard to track the evolution of an event. A story chain is composed of a set of news articles that reveal hidden relationships among different events. Traditional keyword-based search engines provide limited support for finding story chains. In this paper, we propose a random walk based algorithm to find story chains. When breaking news happens, many media outlets report the same event. We have two pruning mechanisms in the algorithm to automatically exclude redundant articles from the story chain and to ensure efficiency of the algorithm. Experimental results show that our proposed algorithm can generate coherent story chains without redundancy.

Mixed Multi-Model Semantic Interaction for Graph-based Narrative Visualizations

“…However, this definition of coherence leads to issues when dealing with data sets that span longer periods of time, such as generating long-ranging connections between events that are seemingly related but too temporally distant to be relevant. To deal with this, previous approaches have used exponential decay factors based on the temporal distance between events [13,42,43,69,72]. Thus, we extend the previous notion of coherence with a temporal component that penalizes events that are temporally distant using an exponential factor.…”

Knowledge-Embedded Narrative Construction from Open Source Intelligence

“…The Thirty-Seventh AAAI Conference on Artificial Intelligence the progression of plot points in events with little to no information gaps. Current narrative construction methods do not provide support for ordering based on contextual features such as thematic frames, and are limited to fill in tasks, given start and end events (Zhu and Oates 2012;Mostafazadeh et al 2017). Constructing narratives based on temporal features alone does not support the dynamic nature of events reported online.…”

Computational Understanding of Narratives: A Survey

“…The sequence of events that construct fragmented narratives can become very detached, due to the difficulties in accurately and efficiently locating evolving themes, concepts, and gaps in ongoing narrative chains [45], [111], [113], [159]. An example of the construction of a fragmented narrative about the January 2022 Texas Synagogue Hostage from multiple online sources is shown in Figure 1.…”

“…As stated previously in Section I, there is a larger focus on computational understanding and extraction of fragmented narratives, which are typically stories divided across collections of newspapers, blogs, and social media posts, and are not statically defined within self-contained stories. A fragmented narrative is a construction of stories where the narration itself is inherently dynamic, constantly shifting the story events as well as the intended outcome [159].…”

“…Zhu et al[159] describe an example of a system that tackles818 the research problem of story completion for fragmented 819 narratives across a variety of news articles. An example of 820 the story chain output is provided in Figure 4.…”