An Efficient Query Retrieval Model Using Pattern Tree-Based Latent Dirichelt Allocation

Today with the advancement in Internet diversified retrieval is
becoming very vast and user want to retrieve the exact
information from many e-commerce sites, e.g. Flipkart,
Snapdeal and so on. Certain number of times, the queries
issued by the user and returned by the search engine is not
entirely similar. Many prior works has been conducted in the
area of information retrieval, however, optimization and time
to retrieve the query still remains a significant challenge to be
addressed. In this work, an efficient Pattern Tree-based Latent
Dirichlet Allocation (PT-LDA) is designed with the objective
of reducing the query retrieval time and therefore improving
the query retrieval ratio. The Topic-based Pattern Tree in PTLDA initially constructs a topic based pattern model using a
relevance factor aiming at reducing the query retrieval time.
Next, Probability-based Latent Dirichlet Allocation performs
an efficient mapping using probability measure aiming at
improving the query retrieval ratio. The Pattern Tree-based
LDA algorithm in PT-LDA framework efficiently maps the
association between user queries and web related documents
in a significant manner. The efficiency of PT-LDA framework
is measured using the Reuters-21578 Text Categorization
Collection Data Set from UCI repository and is shown to be
significantly more efficient in terms of query retrieval time,query retrieval ratio and accuracy of results being obtained when compared to the state-of-the-art works.