PRIS at TREC2013 Knowledge Base Acceleration Track

This paper details the participation of Pattern Recognition and Intelligent System lab of BUPT in CCR and SSF task of TREC 2013 Knowledge Base Acceleration track. In the CCR task, The PRIS system focuses attention on query expansion and similarity calculation. The system uses DBpedia as external source data to do query expansion and generates directional documents to calculate similarities with candidate worth citing documents. In the SSF task, The PRIS system utilizes a pattern learning method to do relation extraction and slot filling. Patterns of regular slots which are same to TAC-KBP slots are learned from KBP Slot Filling corpus. Other slots are filled by following some generalized patterns learned from external source data including homepages of some famous people and facilities. Experiments show that the CCR system gives a good performance above the median value. The pattern learning method for SSF task gives an outstanding performance.