Dynamic Document Annotation for Efficient Data Retrieval

Abstract

Document annotation is considered as one of the most popular methods, where metadata present in document is used to search documents from a large text documents database. Few application domains such as scientific networks, blogs share information in a large amount is usually in unstructured data text documents. Manual annotation of each document becomes a tedious task. Annotations facilitate the task of finding the document topic and assist the reader to quickly overview and understand document. Dynamic document annotation provides a solution to such type of problems. Dynamic annotation of documents is generally considered as a semi-supervised learning task. The documents are dynamically assigned to one of a set of predefined classes based on the features extracted from their textual content. This paper proposes survey on Collaborative Adaptive Data sharing platform (CADS) for document annotation and use of query workload to direct the annotation process. A key novelty of CADS is that it learns with time the most important data attributes of the application, and uses this knowledge to guide the data insertion and querying.