SMU-SIS at TAC 2010 - KBP Track Entity Linking

• Main Authors: GOTTIPATI, Swapna, JIANG, Jing
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2010
• Subjects: Databases and Information Systems
• Online Access: https://ink.library.smu.edu.sg/sis_research/1361; https://ink.library.smu.edu.sg/context/sis_research/article/2360/viewcontent/SMU_SIS_at_TAC_2010___KBP_Track_Entity_Linking.pdf
• Institution: Singapore Management University

Entity linking task is a process of linking the named entity within the unstructured text to the entity in the Knowledge Base. Entity liking to the relevant knowledge is useful in various information extraction and natural language processing applications that improve the user experiences such as search, summarization and so on. We propose the two way entity linking approach to reformulate query, disambiguate the entity and link to the relevant KB repository. This paper describes the details of our participation in TAC 2010 - Knowledge Base Population track. We provided an innovative approach to disambiguate the entity by query reformulation using the query context and Wikipedia knowledge through heuristic approach. We participated in Entity-Linking task using KB text and Entity-Linking task without using KB text tasks. We developed several entity linking engines to evaluate our solution. We compared our methods with a baseline approach and analyzed the experimental results. For both the tasks our system performance is competitive with 76% and 75% mean average scores respectively.