You are here

A Framework for Knowledge Discovery in the News Media Using Text Mining Technique

Journal Name:

Publication Year:

Author NameUniversity of Author
Abstract (2. Language): 
The increase in the media information available to the news audience is growing at an alarming rate and this calls for an efficient means of analyzing and mining information that can be used in making vital decisions. This is most especially true in Nigeria where there is an alarming increase in the crime rate that needs to be analyzed. The work aimed at the use of text mining techniques for deducing the rate of occurrence of crime in Nigeria cities. A review of various text mining techniques with their comparative advantages is presented. The work modeled a text mining framework for the analysis, extraction and organization of news media objects. Mathematical and algorithmic text mining techniques were applied in the development of the crime analysis system. Information was captured from online news material and on ontology was built for representation of specific domain of crime using frames knowledge representation language and Protégé ontology builder. The k-means algorithm is used for knowledge discovery to extract relevant information from the news corpus. The applied clustering technique showcased resulted patterns with graphical representation that aids deductions of the rate of occurrence of crimes. In conclusion, the approach is found useful for sociological analysis of occurrence of crime which invariably aids decision makers on national security.
201
205