Text Document Annotation and Retrieval Based on Content of the Document and Query Workload

Performing search and retrieval in large collection of textual data is a complex task. For effective searching of text documents, annotations are used. Annotations are the tags, s, attribute-value pairs, comments or summary that are attached to a document or a part of the document. Annotations can be considered as a structured representation of unstructured data. Since manually annotating each document in a large collection is not feasible, automatic annotation techniques are used. In this work, an automatic annotation generation technique based on the content of the document and query workload is introduced. Annotations are generated in the form of attribute-value pairs as they are more expressive than simple annotations. The system generates both attributes and values for a document by analyzing the content of the document, annotations of the existing documents and query workload. These annotations are later used during the search and retrieval process for matching with the queries given by the user. As an enhancement to the system, a new ranking method for ranking the retrieved documents is also introduced.