Ranking of query results
It is one of the basic problem in information retrieval.The basic idea behind this problem is given a query and a data set that matches the query,sort the result using some parameters.
There are basically two ranking based models:-
1) Query Dependent Model
2) Query Independent Model
Query Dependent Model=>
First came the Boolean model that can only predict whether a document is relevant to the query or not, but cannot predict the degree of relevance.To improve this degree, the VSM (Vector Space model) was came.Both documents and queries are represented as vectors in a Euclidean space.TF-IDF weighting has been used to get effective results.
After that came BM25 . It is to rank documents by the log-odds of their relevance. BM25 is not a single model, but it de nes a whole family of ranking models, with slightly different components and parameters.
Query Independent Model=>PageRank is an example of this model because it is based on its own importance.
where,U=number of outlinks
N=total pages on web
(1-a)=damping factor
There are basically two ranking based models:-
1) Query Dependent Model
2) Query Independent Model
Query Dependent Model=>
First came the Boolean model that can only predict whether a document is relevant to the query or not, but cannot predict the degree of relevance.To improve this degree, the VSM (Vector Space model) was came.Both documents and queries are represented as vectors in a Euclidean space.TF-IDF weighting has been used to get effective results.
After that came BM25 . It is to rank documents by the log-odds of their relevance. BM25 is not a single model, but it de nes a whole family of ranking models, with slightly different components and parameters.
Query Independent Model=>PageRank is an example of this model because it is based on its own importance.
where,U=number of outlinks
N=total pages on web
(1-a)=damping factor
Comments
Post a Comment