engleski

Identification of persons and business subjects in text documents based on lexical analysis and scoring system

The amount of text documents and textual media news that is created every day on the Internet is growing rapidly, making it very difficult to find useful information effectively. The paper presents a system that identifies persons and business subjects in newly published text documents and matches them with persons and businesses previously stored in a database. The implemented system employs lexical analysis and scoring algorithm tagging the input documents with subjects' id from the database and enabling easy and effective search. Consequently, only the search object and the surrounding context is displayed to the end user. The system is currently successfully used in a web portal.

data mining; information retrieval; lexical analysis

nije evidentirano