StoryBlink: a semantic web approach for linking stories

Finding relevant content automatically is not straightforward due to the unstructured nature of large text corpora. Moreover, traditional techniques to extract structured information out of these corpora are mostly very fine-grained, which deteriorates the needed high-level overview to compare publications. Also, publishing this information as Linked Data can provide for very important context information. This demo paper describes “StoryBlink”, a Web application that enables the discovery of stories through linked books. By finding paths between stories that are represented by a compact set of semantic concepts, it provides the user with relevant stories, based on previously selected publications. By also returning the common concepts between these stories, it gives the user a quick insight into how certain stories are connected. As such, StoryBlink enables an automatic content-based discovery journey between linked stories.