Knowledge Discovery in Online Repositories: A Text Mining Approach

Before the advent of the Internet, the newspapers were the prominent instrument of mobilization for independence and political struggles. Since independence in Nigeria, the political class has adopted newspapers as a medium of political competition and communication. Consequently, most political information exists in unstructured form and hence the need to tap into it using text mining algorithm. This paper implements a text mining algorithm on some unstructured data format in some newspapers. The algorithm involves the following natural language processing techniques: tokenization, text filtering and refinement. As a follow-up to the natural language techniques, association rule mining technique of data mining is used to extract knowledge using the modified Generating Association Rules based on Weighting scheme (GARW).