Abstract
The final phase is classifying new documents on the basis of association-based classifier rules (ABCR), which are obtained from the previous phase. The keys to construct the newly document vector is the use of the like HTML parser and string token that are used in the index level. Specific, the parser is used to compute the general text and term-ranked “important” text. The string tokenizer process is used to parse the terms from the document. Then, the “useful” terms are placed into a table to allow their frequencies newly document to be computed. same heuristic for the term “usefulness” is used in the indexing to find the document frequency of each term. Thus, the tf.idf values for each term in the newly document vector can be computed. ABCR are organized in descending order of support. When a new document is classified, the first matching rule in the sequence conducts the prediction, which is the OABCR. This sequence is usually tied by a class. When no rule in the sequence matches the coming document, the class of the document is predicted by scanning other class label rules. If no class rule matches the document attributes, then the document class label becomes null.
Keywords
Subject Areas
Citations by Year
Related YTU Activities
Sustainability activities and projects connected to this article.
