Abstract. In this paper, we present a method for the automatic extraction of numerical fields (zip codes, phone numbers, etc.) from incoming mail documents. The approach is based o...
In this manuscript we present the summarization and categorization subsystems of a complete mechanism that begins with web-page fetching and concludes with representation of the c...
Current Data Mining techniques usually do not have a mechanism to automatically infer semantic features inherent in the data being “mined”. The semantics are either injected i...
A challenging problem in open information extraction and text mining is the learning of the selectional restrictions of semantic relations. We propose a minimally supervised boots...
Semantic orientation (SO) for texts is often determined on the basis of the positive or negative polarity, or sentiment, found in the text. Polarity is typically extracted using th...