An Unsupervised Web Data Extraction System

Abstract

The web has become a large collection of many unstructured data or documents. To search for useful information on the web, search engines are to be used. Even from the retrieved results, users are required to search within those documents to Find informa- tion. So it has become difficult to extract the information easily. To solve this issue, use of different web data extraction techniques is to be done. To achieve this goal, few unsupervised web data extraction systems have been studied and there are many di er- ent techniques available for information extraction. A survey for the existing Roadrunner Algorithm which is effcient for data extraction is done. Still, it has few limitations and to overcome that, an approach is proposed which uses Mining Data Records and Tree Align- ment technique for processing of the input HTML pages. An experiment is performed to compare both the results and it is able to overcome the limitations of Roadrunner. So, it can be used to extract the useful data and get the desired results.