REDUCING REPLICA OF USER QUERY CLUSTER- CONTENT AND SUB-HYPERLINKS IN THE SEARCH ENGINE LOG BASED USER PROFILE
P. Srinivasan, Krishnan Batri · 2013
Most important visible component of the internet contains millions of web pages waiting to present information on an amazing collection of topics. The search engine has played an increasingly important role. Nowadays, the search engine based string matching has inbuilt troubles like a low accuracy, inadequate individual support, duplicates document and replica of hyperlinks in the user profile and so on. This work introduced a method against previously proposed personalized query clustering method by other authors. Experimental results show that a profile captures and utilizes both user’s replica and nonreplicated cluster-content and links. The non replica of hyper, sub-hyperlinks and cluster-content perform the best of among results. An important result from the experiments is that profiles with replica of links can increase the separation between similar and different queries. The separation provides a clear threshold value for a LINGO clustering algorithm to terminate and improve the overall quality of the resulting query clusters and hyper, sub-hyper links to improve search engine performance through user profile.