Context-Aware Visualization of Entity-Entity Relationships in a Document Corpus
The paper presents an interactive graphical environment, which enables the detection and graphical visualization of concepts in a document or a document collection. Concepts are expressed by (multiple) entities extracted from the document and by the relationships between them. The tool offers an entity-centered view, which graphically shows the most important relationships of a central entity or entity-group, consisting of multiple co-occurring entities. By specifying prefixes and an additional available type system, complex filters can be created that allow the disclosure of various relationships between entities. Entities and their relationships are determined at the time of indexing and stored in appropriate data structures, so that an interactive search and exploration of relationships between entities is easily possible at runtime. The tool is available for online demonstration at https://www.smiffy.de/CoOcViz (credentials: adc2019, password: demo) .
KeywordsNamed entities Document semantic Entity co-occurrence Visualization Navigation
- 2.Hoffart, J.: Discovering and disambiguating named entities in text, Ph.D. dissertation. Universität des Saarlandes, Saarbrücken (2015)Google Scholar
- 3.Schmidt, A., Hoffart, J., Milchevski, D., Weikum, G.: Context-sensitive auto-completion for searching with entities and categories. In: Proceedings of the 39th International ACM SIGIR Conference, pp. 1097–1100 (2016)Google Scholar
- 4.Hoffart, J., Milchevski, D., Weikum, G.: STICS: searching with strings, things, and cats. In: Proceedings of the 37th International ACM SIGIR Conference, pp. 1247–1248 (2014)Google Scholar
- 5.Seufert, S., Berberich, K., Bedathur, S.J., Kondreddi S.K., Ernst, P., Weikum, G.: Espresso: explaining relationships between entity sets. In: Proceedings of the 25th ACM International CIKM Conference, pp. 1311–1320 (2016)Google Scholar