Espresso: Explaining Relationships between Entity Sets
Espresso is a system to compute semantically meaningful substructures (so-called relatedness cores) from a knowledge graph. The purpose of the system is to answer questions of the form «Which European politicians are related to politicians in the United States and how?» or «How can one summarize the relationship between China and countries from the Middle East over the last five years?» In this setting, a question is specified by means of two sets of query entities. These sets (e.g. "European politicians" or "United States politicians") can be determined by an initial graph query over a knowledge graph capturing relationships between real-world entities. As a next step, we analyze the (indirect) relationships that connect entities from both sets (e. g. membership in organizations, statements made on TV, etc.), generate an informative and concise result, and finally provide a user-friendly explanation of the answer. As output, we aim to return concise subgraphs corresponding to important event complexes, that connect entities from the two sets and explain their relationships. Espresso provides a user interface for the specification of entity sets, computes informative relatedness cores that summarize the relationship between the query entities, and finally displays a visually appealing visualization of the extracted subgraph to the user. Applications of the proposed system include scenarios that require to provide background information on the current state-of-affairs between real-world entities such as politicians, organizations, and the like, e. g. to a journalist preparing an article involving the entities of interest.
- Espresso: Explaining Relationships between Entity Sets
Stephan Seufert, Klaus Berberich, Srikanta J. Bedathur, Sarath Kumar Kondreddi, Patrick Ernst, and Gerhard Weikum
Proceedings of the 25th International Conference on Information and Knowledge Management (CIKM 2016),
Indianapolis, IN, United States, October 24-28, 2016. ACM.
- Instant Espresso: Interactive Analysis of Relationships in Knowledge Graphs (Demo)
Stephan Seufert, Patrick Ernst, Srikanta J. Bedathur, Sarath Kumar Kondreddi, Klaus Berberich, and Gerhard Weikum
Proceedings of the 25th International World Wide Web Conference (WWW 2016),
Montreal, QC, Canada, April 11-15, 2016. ACM.
- Efficient Computation of Relationship-Centrality in Large Entity-Relationship Graphs (poster)
Stephan Seufert, Srikanta J. Bedathur, Johannes Hoffart, Andrey Gubichev, and Klaus Berberisch
Posters and Demonstrations Track of the 12th International Semantic Web Conference (ISWC 2013),
Sydney, NSW, Australia, October 21-25, 2013.
|ClueWeb12 Cooccurrence||Entity-cooccurences in ClueWeb||entity1,entity2,count||Download||728M|
|Entity||Collection of all entities||id, yagoid, freebaseid, wpid, name, readable yagoid, event (t/f)||Download||117M|
|YAGO Types||Collection of YAGO types||id,name||Download||4.6M|
|Entity-YAGO Type||Association of entity with YAGO type||entity,type||Download||169M|
|Freebase Types||Collection of Freebase types||id,name||Download||64K|
|Entity-Freebase Type||Association of entity with Freebase type||entity,type||Download||29M|
|Links||Links between entities||source,target,MW-similarity,KORE-similarity||Download||798M|
|Relations||Collection of relations||id,name,count||Download||423|
|Link-Relations||Association of links with relations||source,target,relation||Download||259M|
|Popularity||Entity popularities based on pageviews||entity,pop||Download||34M|
|Views||Pageviews for entities||entity,day,count,Z-score,relative popularity||Download||32G|
|Snippets||Short textual entity descriptions from Wikipedia||entity,snippet||Download||462M|
|ClueWeb12 Counts||Number of entity occurrences in ClueWeb||entity,count||Download||8.4M|
Provided files are BZ2-compressed CSV files with header and double quoting.
- The datasets contain material from Wikipedia, which is released under the Creative Commons Attribution-Share-Alike License 3.0.
- The datasets contain data derived from ClueWeb12.
- The datasets contain material from Freebase.
- The datasets contain material from the Wikipedia Pageview project.