Automatic annotating SRRs from web databases using Naive Bayes approach

The number of web databases are increasing progressively day-by-day and are web accessible through HTML- based form. The data units that are encoded in the search result web page are in a particular structure and unstructured format and are needed for human browsing for applications like, comparison shopping, to rate a web resource, deep web collection etc. So, for machine processing, it is needed to extract all data units at one place and assign semantic labels accurately to the retrieved data units. In this paper, an automatic annotation system is introduced in which the data units from SRRs are extracted, automatic semantic labels are obtained from the data units extracted and the values of the attributes are aligned accurately under the semantic label. To improve the automatic generation of annotations, Naive Bayes machine learning classifier is used. Our results obtained shows that use of proposed system contributes to achieve accurate results.