Download now Free registration required
Text mining based on the integration of Information Extraction (IE) and traditional Knowledge Discovery from Databases (KDD). The authors, first present the idea of combining IE and KDD serially for text mining, explain how a document in this system can be represented as a vector of textual elements, and empirically show that rules mined from IE-extracted data are nearly as accurate as those discovered from manually extracted data. The assumption of traditional data mining that the information to be mined is already in the form of a relational database does not hold in many cases.
- Format: PDF
- Size: 201.86 KB