Measuring the Impact of Using Different Tools on Classification System Results

Abstract A huge amount of textual data is available on the web. These data need to be classified under labels or classes to make the search more efficient and easier. Achieved by using automatic classification is used for this task. Many factors impact on the performance of the classifier system, such as the amount of using dataset, the data dispersion degree, preprocessing tools, feature extraction methods, terms weighting, and data reduction. So, researchers constantly compete to build a robust classifier with good performance. This study focuses on the effect of using different tools in preprocessing and term weighting stages. The experimental results applied on two different languages (Arabic and English languages). Also, the experimental results were compared with the recent related works.