Pattern Discovery for Text Mining Measured by Levenshtein Edit Distance

The text mining techniques are utilized to extract interesting knowledge or information from the text documents. The process of finding accurate knowledge in text documents represents the main challenge for the users. Since many existing techniques of text mining encompassed term-based approaches, these techniques face the issues of synonymy and polysemy. Over the years, the researchers have discovered the assumption that pattern-based methodologies must provide better performance than the term based ones. This paper proposes an effective and innovative technique for pattern discovery that contains the pattern taxonomy model and Levenshtein edit distance algorithms for improving the efficiency of utilizing and updating the obtained patterns to find interesting and relevant information. This system shows that the accuracy of PTM is 98.09% for a specific dataset.