i need ppt on ranking on data manifold with sinkpoints for seminar please mail me at ravoorsangamesh[at]gmail.com

INTRODUCTION
Rating is a very important problem in natural language processing, data mining, and information retrieval.
As a rule, rating the problem is explained as follows: given a set of data objects, the ranking function organizes the objects
in the set according to their level of importance, relevance or preferences. But the mass of documents takes Might
that contain duplicate and redundant information is not very useful to users. This redundancy is in the top rank
The results reduce the chance of a multifaceted to meet the needs of different users. Thus, diversity in the ranked results are very
important in addition to the relevance and importance in the ranking of documents. Diversity in the ranking of documents helps users
to handle user requests with explicit queries relating to the information contained in the documents stored online. The variety is
the means to get the relevant information very easily.
Many real-time applications require diversity in approaches. In addition to the relevance and importance,
Also considered to be a variety as an important Criterion in the extraction of data a year. For example, in recommendation requests, requests
This is a recommended rating approaches should capture the different requirements of different user query (that is) the
We also recommend that you request must be different. The concept of diversity in the application of the generalization, the brief
Sentences of the candidate must be smaller and should also contain various aspects of information relevant to the request.The
The concept of diversity is widely studied in recent years. Many of them have suggested that a variety of approaches to
Such as many domains on the basis of cluster centers of gravity, maximum relevance and diversity section. But these
Approaches rarely consider the relevance and diversity in a single order.
Diversity rating process makes use of a variety of dive points. This indicates that sink woven data objects
contain a minimum of fixed rank. Ranking assess data objects closer to the sink has Naturally
What is included in the ranking process is usually based on the inside of the collector. Thus, diversity and value as
as well as the relevance of diversity captured during the process of ranking.