Rule-Based Entity Resolution on Database with Hidden Temporal Information

摘要: In this paper, we deal with the problem of rule-based entity resolution on imprecise temporal data. Entity (ER) is widely explored in research community, but data, especially without available timestamps, has not been studied well yet. Because elapsing time, records referring to same observed different time periods may be different. Besides traditional similarity-based ER approaches, by carefully exploring several data quality rules, e.g., matching dependency and currency, much information can obtained facilitate cope problem. use such rules derive records’ order trend their attributes’ evolvement time. Specifically, first block into smaller blocks, then currency constraints, propose a clustering approach two steps, i.e., skeleton banding clustering. Experimental results both real synthetic show that our method achieve high accuracy efficiency datasets hidden information.