Data extraction which is important for many applications extracts the record from HTML files. The existing Machine learning method requires human labeling of web sites for extracting data from web pages. So, this process is very time consuming. Automatic pattern discovery method enables inaccurate alignment of multiple data records in web pages. Taken into consideration the limitations of data extraction methods. Many applications necessitate the automatic extraction of data from the query result pages. The proposed method of data extraction using Web extraction tool automatically extracts data from query result pages. The data extracted using web extraction tool is aligned in a structured format using Cosine-Similarity