Joint inference for cross-document information extraction

Previous information extraction (IE) systems are typically organized as a pipeline architecture of separated stages which make independent local decisions. When the data grows beyond some certain size, the extracted facts become inter-dependent and thus we can take advantage of information redundancy to conduct reasoning across documents and improve the performance of IE. We describe a joint inference approach based on information network structure to conduct cross-fact reasoning with an integer linear programming framework. Without using any additional labeled data this new method obtained 13.7%-24.4% user browsing cost reduction over a state-of-the-art IE system which extracts various types of facts independently. Categories and Subject Descriptors I.2.7 [Natural Language Processing]: Text analysis; H.3.4 [Systems and Software]: Information networks, Question-answering (fact retrieval) systems General Terms Algorithms Keywords Information extraction, global reasoning, Integer Linear...

• Type: Journal
• Year: 2011
• Where: CIKM
• Authors: Qi Li, Sam Anzaroot, Wen-Pin Lin, Xiang Li, Heng Ji
• Keywords: CIKM 2011, Information Redundancy, Information Technology, Natural Language Processing, Pipeline Architecture