II. RELATED WORK
This section provides review of literature on document streams and making top-k queries in such streams. For effective document retrieval different filtering techniques are explored in 1, 2 and 3. Content matching approach for document retrieval in the domain of publisher/subscriber is investigated in 4. On the other hand top-k matching in the systems that are based on publisher/subscriber is explored in 5. Finding k most relevant publications in the publisher/subscriber model is studied in 6. With respect to social annotation based news, in the context of publisher/subscriber, a system is proposed in 7 for top-k publishing. Many top-k query processing techniques are discussed in 8. Sometimes, it is better to aggregate results so as to make it more meaningful. Such work is done in 9 for making middleware software more useful. Working with multiple aggregations with respect to streaming documents is made in 21.
Text mining and performing inverted search operations is the main focus of the study made in 10. There are two-level procedures followed in 11 for evaluating queries processing. With respect to top-k queries, evaluation procedures are discussed in 12. With respect to sliding windows, the problem of continuous top-k monitoring is explored in 13. Top-k queries with user preferences is investigated in 14 for continuous performance with respect to top-k queries. Evaluation of such continuous top-k queries is made in 15 and 16. With respect to document streaming and personalized query processing is investigated in 17 by using web 2.0 streams. Monitoring of personalized hot news in the context of web 2.0 document streams is made in 18. Documents with non-homogenous scoring functions are studied for processing continuous queries is made in 19 while approximate top-n queries on the documents streams are the study in 20. In 22 a technique known as reverse identifier ordering is employed and they improved it to have efficient top-k queries and monitoring on document streams. From the review of literature it is understood that there has been considerable research on top-k queries on document streams. In this paper we found the suitability of adaptive approach to improve it further by considering it as an optimization problem.
II. RELATED WORK