Web Database Annotation for Fast and Accurate Retrieval

Mounting of databases have become web accessible through Hyper Text Mark up Language based explore interfaces. When people browse for particular purpose the search results are obtained relevant to data or irrelevant to data. To get the desired data people need to search again from the obtained search results. To overcome this problem, we introduce an automatic marginal note method that initially arranges the data units on a resultant page of a search site into different groups. Then, for each and every group we build marginal notes from various aspects and combine the distinct labels to estimate an absolute label for it. Lastly marginal note binder for each site is dynamically constructed and used for making annotations automatically when we search newly from the same web DB.