作者: Raymond Vermaas , Damir Vandic , Flavius Frasincar
DOI: 10.1007/978-3-319-11749-2_40
关键词:
摘要: Product entity resolution is an important part of online product search, where entities coming from different websites need to be aggregated in the search results. In this paper, we propose approach using descriptive power ontology. our algorithm, use similarity measures that are defined specifically for each type feature and learn weights by means a genetic algorithm. evaluation obtain F 1-measures 59% 72% two classes consider. The obtained results significantly better than those state-of-the-art