作者: Sundararajan Sellamanickam , Mridul Muralidharan , Philip L. Bohannon , Ashwin Machanavajjhala , Sathiya K. Selvaraj
DOI:
关键词: Feature (machine learning) 、 Identifier 、 Weak entity 、 Web page 、 Home page 、 Rank (computer programming) 、 Class (set theory) 、 Set (abstract data type) 、 Information retrieval 、 Computer science
摘要: A system and method is described for large scale entity-specific classification of each set candidates in a collection specific entity entities. The entities may comprise category or domain (e.g. schools, restaurants, manufacturers, products, events, people). Candidates webpages other resources with resource identifiers. Entity sets be found by leveraging search engine query results user interaction therewith queries based on attributes. relationship(s) class(es) which candidate are being classified relative to an authoritative, official home page (OHP), class fan page, review, aggregator) entity. feature generator generates features candidates. In accordance its features, one more classifiers rank