ABSTRACTIn today’s world,Keywordsuggestion in web search is a very important feature to be considered. The mainproblem for search engine is that the queries provided by user are sometimevery short and short queries are ambiguous.Then result provided by searchengine do not satisfy user’s requirements.User donot know how to express in theform of query or he may have less knowledge about the thing that he wants tosearch.To solve his problem search engine provide a set of keywords as querysuggestion. To increase the experienceof user’s search engine query suggestion is provided. Query suggestion helpsuser to specify query in the form of setof keywords and provide appropriate result to the user, But today’s querysuggestion do not consider location of user and the distance factor betweenuser location and document’s location.
If location of user is considered forquery suggestion then user will get more relevant document. The location awarekeyword query suggestion (LKS) takes the location of user and retrieve thedocument to the user which are close to user’s location. In this paper we willdiscuss the LKS framework and discuss the research work that has been donepreviously. General Terms Bipartite graph,Document proximity,Location-basedsearch,KeywordsLocation-based search, Keyword search, NearestneighborsSpatial databases 1.
INTRODUCTIONSearchengine is a tool to find the information.However many a times user unable tolocate required document to their problems.This process is also known as”struggling search”.In struggling search to achieve required result userstruggles a lot to formulate query. For example user wants to search “apple” then this single word query is achallenge for search engine to provide relevant document for query .In thiscase search engine face the problem to show result as Apple as fruit or Appleas company.
Sometimeuser also looking for information on a topic which is new for user as per hisneeds hence he also faces problem to construct appropriate query. And there arebillions of web pages and user have to search from that information. If therelevant document for user’s query are very deep in result list then at thatpoint user’s query will fail.
Keyword suggestion has become a very important feature of today’sWeb search engines. Query suggestion’s aim is to refine user’s experience ofsearch engine by suggesting alternate query for his first query so that he canget more information and get more relevant document.Many times even literateuser of web faces the problem to formulate a query. After submitting a query, ifexpected result were not retrieved thenuser may not be satisfied with theresults as he was expecting some other results. Major problem of current searchengine is that the queries are short. The biggest problem for search engine isshort query. Sometime user have very less knowledge about topic hence usersubmit short queries.
To provide relevant result for such query search engineshould be very smart to understand user’s requirement.Query suggestion helpsuser to construct a query. Many query suggestion technique has proved that theyfulfill user’s needs. But set of keywords provided by query suggestion is not a gooddescriptor of user’s needs. None of query suggestion methods are locationbased. User’s location is not considered during query suggestion .
The locationaware keyword query suggestion (LKS) providesresult which are relevant touser’s need also located near user’s location. Thelocation aware keyword suggestion (LKS) provide the queries and retrieve documents which are near to user’slocation and at the same time relevant to his needs 2. LITERATURE SURVEYA lot of research work is done on query suggestionpreviously.1PathRank:A Novel Node Ranking Measure on a Heterogeneous Graph for Recommender Systems:In this paper path rank algorithm is used.thepath rank is defined for hetrogenous graph by providing extension to thepersonalized page rank. In path rank,it will make full use and drive benefitsfrom semantics which is present behind the various types of nodes and edges ofhetrogenous graph and follow semantics of content based filtering andcollaborative filtering. 2QueryRecommendation using Query Logs in Search Engines:In thisquery recommender algorithm is used . In this algorithm similarity of query andsupport of query is calculated .
Based on similarity of query the clustering ofquery is done.The clustering process uses the content of historical search of usersregisteredin the query log of the searchengine. The method discovers the relatedqueries and will rank them as per relevance of the query 3context awareIn this clustering queries and query suggestionalgorithm is used.This technique take into account the immediatelyprevious query which has been fired by the user as a context in providing querysuggestion.It will correlate pervious query to the new query and try to providequery suggestion.
4 Agglomerative clustering of a search enginequery log :In this graph based iterative clustering techniqueis used in which clustering of user’s query and url is done.It will mine the user’s transaction with searchengine.and make the clusters of similar queries and similar urls.5time aware:In this time aware query suggestion,query clustering,queryselection technique is used.
For query clustering TA-clustring algorithm isused.which clusters query suggestions with timeline so that the user can findout his search from a temporal aspect.furthermore, when an suggested query will be clicked, TaSQS will provide related web pagesfrom query-URL bipartite graphs after providing ranking to them based on the click counts within a specific time period.6 QuerySuggestion Using Hitting Time In this query suggestion using hitting time algorithmand personalized query suggestion technique a time are used for providing querysuggestion using hitting is used.Query suggestion algorithm is based on providingranking to queries along with hitting time.
Without any involvement of twisted or heavy tuning of parameters, this techniqueclearly Presents the semantic consistencybetween the query suggestion provided and the original query.7Spatio textual similarity joins:Ourwork is related to spatial distance joins, set-similarity joins, andspatio-textual search. batch processing technique is used and our methodsexploit spatial indexing and pruning techniques to reduce the space where the(more expensive) textual similarity predicate needs to be verified; for thelatter, they adapt the state-of-the-art algorithm for set similarity joins.8 OptimalRare Query Suggestion With Implicit User FeedbackIn this paper,Construct two query-URL bipartite graphs fromquery logs, where the click graph contains query-URL click information and the skipgraph contains query-URL skip information, perform random walk on each of thegraphs, using the random walk with restart (RWR) technique Build a correlation matrix for URLs from the category ofURLs,based on the URL correlation, iteratively optimize the model to estimatethe best parameters of random walk and the combination rate of click and skipgraphs.