抽象的

Efficient Ranked Keyword Search Using AME

K.Padmapriya, A.Ambika , A.Gayathiri

Entity Recognition is process of identifying predefined entities such as person names, products, or locations in a given document. This is done by finding all possible substrings from a document that match any reference in the given entity dictionary. Approximate Membership Extraction (AME) method was used for finding all substrings in a given document that can approximately match any clean references but it generates many redundant matched substrings because of approximation (rough calculation), thus rendering AME is not suitable for real-world tasks based on entity extraction. We propose a web-based join framework which combines a web search along with the approximate membership localization. Our process first provides a top n number of documents fetched from the web using a general search using the given query and then approximate membership localization(AML) is applied on these documents using the clear reference table and extracts the entities form the document to form the intermediate reference table using Edit distance Vector, Score Correlation.