Efficient way of user search location in query processing

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
__________________________________________________________________________________________
Volume: 03 Special Issue: 07 | May-2014, Available @ http://www.ijret.org 564
EFFICIENT WAY OF USER SEARCH LOCATION IN QUERY
PROCESSING
Parimala S1
, Jayanthi S2
1
PG Student, Department of Computer Science and Engineering, Anna University (BIT Campus), Trichy, India
2
Assistant Professor, Department of Computer Science and Engineering, Anna University (BIT Campus), Trichy, India
Abstract
Rapid growth of the mobile search location information is an important concern. Usually mobile users search the query in service
providers. In this paper, apply click through for finding the users’ interest in search engine mainly from their own search database.
User preferences are classified as content concept and location concept. Ontology concepts are used to store the user preferences in
the client side and Content extraction, Re-ranking are used in the server side to get the closest point search results. In this, use client–
server architecture to get the user relevant information. Server provides the result based on user search. Classify the query based on
the user click through. Instead of producing global search result, it performs ranking to get closest point result based on the server
database. Frequently accessed user queries are stored in the client side for fast access. If the result is not available, then the user
queries can be posted to get the updated result when the server is updated.
Keywords: content concept, location concept, mobile search, user search.
---------------------------------------------------------------------***----------------------------------------------------------------------
1. INTRODUCTION
In web search create many challenges. Because whenever
different user enters the same query it displays the same result.
For example whenever the user enter the query it displays
different kind of information like Apple query have computer
and fruit kind of information. In normal web search must
specifies the user interest with query. In web search enter the
query and select the hierarchy before submit the query.
Difficult to find the correct link before submit the query.
In personalised never use general profile information for filter
the user profile. Personalised search is a experience of the
user’s query. It contains the history and reranking search
results. Instead of use general search engine, use personalised
search make information more secure with unique interest.
Another user couldn’t find which data search by user.
Personalised search is classified the user different information
based on their search result and improve the search quality.
Even though have GOOGLE personalised search get personal
information in public link without user permission. Whenever
create user account provide access their personal information
to the server. Get the information from the web done by search
engine.
In GOOGLE search problem is user and search engine
interactions are less. It must confine the user’s click through
for analyze about the user’s interest Example whenever the
user need to know the hotels in Japan. He just enter hotel in
Japan based on the query classify hotel is content information
and Japan is location information. Personalised search use
client and server get the exact result with efficient.
In this paper, user is used for classify the location and content
information based on the user’s click through and server is
used for provide the information based on the classification
and ranking the results. Find the how much like the link based
on the click for provide effectiveness to the personalised
search must classified link by content and location.
The content and location preferences are used for the search to
the preferences are used for the search to the user. It can also
like geo query and non-geo query mainly focus on location
information non-geo query intention to the content
information. Backend search engine doesn’t know the user’s
interest history. Client can set the privacy level to the server.
The server never store information of the user’s like more than
some limitation. Ontology is used find the user preference and
filters the interest filter information forward to the server to
find the relevant information.
Whenever the user enters the query, the search system gets the
information from the search engine. The search engine result
is providing to the search system. The search system contains
the user interest and history and click through data. Search
system provides the relevant information result to the user.
Search engine provide the lot of information but search system
ranking the information based n the user interest. Privacy is
important concern for personalised whenever the user enter the
query like hotel server provide the result as map and room
rate.

__________________________________________________________________________________________
The user interest in roommate based on the user click they
classify as room rate and special discount rate the click
through is stored in client database user. So the information
never stolen by the server. The Server doesn’t know the user’s
interest based on the search find out the important information.
2. RELATED WORK
Mobile search find the user interest by the click through link
of the information. E. Agichtein, E. Brill [1] suggest that Rank
Net is neural net tuning algorithm find out the weight for user
interest. In this based on the thousands of queries provide the
ranking it can understand the human interest by provide their
labels. In the information retrieval use ranking for filter the
information. Ranking improved by implicit feedback.
General search engine results based on the user preferences as
per the ranking of the click through data. Provide the score for
click through data.Provide the high score for click data than
the un-clicked data. In this use vector features for rank the user
preferences and train the features for ranking the function.
Each time the user select provide the score for which link
select for interest, Ranking results based on lot of features, like
content based features, query-independent page quality
features. It contains three tyes click through features and
browsing features and Query text features. Implicit feedback is
used to improve the ranking result of the search engine results.
Y-Y.Chen, T.Suel[2] suggest that query foot print and
geographic foot print for separate the query of the user. Use K-
Sweep algorithm for fetch the information. From the disk the
information is fetch without order the data display to the user.
In this use text index structure for able to find the user interest
document based on the word.
J.Attenberg[3] proposed that query classified as geo-query and
non-geoquery for improve the user search. Whenever user
enter the query filter based on classification of the geographic
term and queries with no location information classified as
non-geo query based AOL trace differentiate geo and non-geo.
In normal search whenever the query have spell mistake and
provide the corrected query. Sacrifice of quality of service
some queries may be dropped.
T. Joachims[4] propose that SVM algorithm use user interest
for ranking the document and whenever click the document it
decide as positive document and then ranking the link.
Whether the document provides negative weights means it’s
not rank.
K.W.T.Leung, D.L.Lee [5] proposed that OMF is capture
location and content concept for get the relevant information
result for the user query. Content concept is keyword in the
link. The location concept is physical location find from the
link. Whenever user enter the query and then select the link
extract the content and location in the OMF profile. Joachim’s
method User can search the query result from top to bottom.
User skip the document to read web snippet understand it
prefer another document s skip it. It is used to mining the
document and rank the result based on the user preferences
user select the document di , but skip the document dj because
of the user interest. dj < r, di . r is the user interest ,user like the
document di. Joachim’s output is input to the ranking process.
K.W.-T. Leung, W. Ng [6] suggest that Whenever the user
enter the query that forward to the middleware then the query
passed to the search engine. Search engine provide result with
the some link to the middleware.
BB (Beeferman and Berger’s )agglomerative clustering
algorithm effective technique in clustering. This algorithm
creates a bipartite graph based on the user query, whenever the
user click the document relation between the query and
document create bipartite graph. Provide cluster algorithm for
the graph to find out the similar query and similar document.
Q. Tan, X. Chai [7] suggest that RSCF is providing efficient
training for even small dataset. RSCF select the log files and
then extract the data then provide classify as labelled data and
unlabelled data. RSCF provide better ranking than the RSVM
algorithm. RSCF create a meta search engine that contain the
MSN search, wisent and overture. In this search mainly focus
on the local location as higher interest than the global location
result. In geographic search query is extracted based on the
city name, address and map by external database.
3. LOCATION SEARCH OVERVIEW
In this paper user search the location of the query in the
mobile environment. mobile user enter the query to the client.
In the client have two database as location database and
content database for store the user information separately.The
client forward the query to the server for get the result. The
server contain the location and content database for retrieve
the data. The user enter the query as university in trichy the
query is passed to the client for check the query information
available if user already search the query then get
information from the client database also separate data also. If
the query is new for client then forward query to server for
that reduce the workload of the server.
In server side contain all information. In the client update the
query by the user whenever user didn’t get the exact result
from the server. In server side use content and location
classification for ranking the location and find the user
interest. The server display the result with nearest location and
suggestion result then find the shortest path for the query of
the user.

__________________________________________________________________________________________
Fig 1 System Architecture
4. CLICK THROUGH USER INTEREST
Learning the user interest based on the user click through data.
User query classified as content classification and location
classification . client have two database for store the
information. Server also contain the location and content . In
the existing system whenever the user enter the query. In the
client-side query classified as content concept and location
concept. Before system user must type manually their interest
mention with query and lot of interest without relevant
information also display to the user.
From Joachim’s Method select the user interest. Result is
search from top to bottom for select the relevant and interest
result of information. problems in that paper User can get the
available data only, even can’t provide the needed information
in database for future database. Ranking is based only by the
user click. Each time client asks information from server even
repeat query itself.
5. USER SEARCH RANKING THE LOCATION
In the user search provide the ontology update in the client
side. Client get the result of query without network connection
from the ontology if the query already search instead of ask to
the server result provide from client database. But each time
the ontology is update whenever the user enter the query, If
doesn’t get the relevant information, then user update provide
commitment to relevant information result to the ontology. As
like existing the query classified as content and location
concept. SpyNB method finds out the user point of interest
location.
SpyNB method extraction only input to the ranking process.
Based on the positive sample provide the ranking and find the
minimum distance location of the query for ranking. In the
server–side also, have content and location concept database
for filter the user query. In the server side also mine the unlike
of the user.
Based on the location of query provide nearest neighbour
location result also with separate link. In this paper use
module as client side search and server side search and
clickthrough data at last use ranking the user interest is based
on the nearest location also. Repeated query result from client
database, don’t need to forward the query server. Ranking the
query result based on Euclidean distance, so find with shortest
path. Nearest five location display whenever user ask the
query location. In the pending database insert the user doesn’t
get result query.
Whenever server get exact result if update in the pending
database to indicate the user. Implement the location search
with client and server side search and the click through data
and ranking the user interest from. In this paper implement the
location search with location and information of the place
whenever user ask query forget result.
6. EXPERIMENTAL RESULT
In this section estimate the closest point search mechanism.
Evaluate location and content query by user profile. Provide
the quality to the user for get the relevant result in less
duration. Verify the user profile for help to improve the search
quality.
Fig 2. location query range
The Figure 2 shows the range of the closest point and the
relevant data based on the results of the user queries.
0
10
20
30
40
50
60
70
80
Range
Location
Closest
Point
Relevant
Data

__________________________________________________________________________________________
7. CONCLUSIONS
In personalized search query processing technique is used for
ranking the query result based on the click through data.
Categorization of content and location concept is easily
identified and classified based on the user interest. In this
paper content and location database are stored in client side.
Our work is to categorize the query and perform ranking in
server side. Our work is to extract the data and find out the
query result. Client submit the request to the server. The
Server solves the problem by avoiding the non-relevant data
and provide relevant information by user click through and
ranking process. The user is provided with the closest point
result from the server.
REFERENCES
[1] E. Agichtein, E. Brill, and S. Dumais, “Improving Web
Search Ranking by Incorporating User Behavior
Information,” Proc. 29th
Ann. Int’l ACM SIGIR Conf.
Research and Development in Information Retrieval
(SIGIR), 2006.
[2] Y.-Y. Chen, T. Suel, and A. Markowetz, “Efficient
Query Processing in Geographic Web Search Engines,”
Proc. Int’l ACM SIGIR Conf. Research and
Development in Information Retrieval (SIGIR), 2006.
[3] Q. Gan, J. Attenberg, A. Markowetz, and T. Suel,
“Analysis of Geographic Queries in a Search Engine
Log,” Proc. First Int’l Workshop Location and the Web
(LocWeb), 2008.
[4] T. Joachims, “Optimizing Search Engines Using
Clickthrough Data,” Proc. ACM SIGKDD Int’l Conf.
Knowledge Discovery and Data Mining, 2002.
[5] K.W.-T. Leung, D.L. Lee, and W.-C. Lee,
“Personalized Web Search with Location Preferences,”
Proc. IEEE Int’l Conf. Data Mining (ICDE), 2010.
[6] K.W.-T. Leung, W. Ng, and D.L. Lee, “Personalized
Concept-Based Clustering of Search Engine Queries,”
IEEE Trans. Knowledge and Data Eng., vol. 20, no. 11,
pp. 1505-1518, Nov. 2008.
[7] Q. Tan, X. Chai, W. Ng, and D. Lee, “Applying Co-
Training to Clickthrough Data for Search Engine
Adaptation,” Proc. Int’l Conf. Database Systems for
Advanced Applications (DASFAA), 2004.
BIOGRAPHIES
Parimala received Bachelor of Technology
in Computer Science Engineering from
Chettinad College of Engineering and
Technology, karur. She is now pursuing her
Master in Engineering, Pervasive
Computing technology in Anna University
(BIT Campus), Trichy. Her areas of interest
are network security, data mining.
Jayanthi received her M.E Degree in in
Computer Science Engineering. She is
now working as Assistant Professor in
Anna University (BIT Campus), Trichy.
Her areas of interest are network security,

Efficient way of user search location in query processing

More Related Content

What's hot (19)

Viewers also liked (20)

Similar to Efficient way of user search location in query processing (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Efficient way of user search location in query processing