DefinePK

Abstract

Active learning tackles data scarcity problem by choosing unlabelled data for labeling and training. Active learning handles large volume of data selection. Data are diverse in character or wide range. There is a problem of handling unlabelled data and certain predefined category. This can be overcome by developing a method which is flexible to handle large volume (diverse in content) are learned through single platform or group of item rather than individually. Performance analysis using data mining approaches validates accuracy and F measure, combines precision and recall and takes data relevant to query that are successfully retrieved and efficiency of active learning leading to reliable and authentic predictions.