Multi-step classification approaches to cumulative citation recommendation
Proceedings of the 10th Conference on Open Research Areas in Information Retrieval
| Hi-index | 0.00 |
In this paper we present KBAAA, a prototype system that provides a web-based interface for the analysis and assessment of knowledge base acceleration systems. KBAAA displays items from a voluminous document stream that are deemed central to a given target entity. Results can be visualised on a timeline along with the number of pageviews and edits made for that entity in the knowledge base. If available, the system also shows relevance assessments and computes evaluation metrics. Further functionality includes side-by-side comparison of two different approaches and free text search over the document collections. The prototype has been fully implemented and deployed on the dataset and target entities of the TREC 2012 KBA track.