29-07-2011, 10:18 AM
Abstract-
Nowadays search engines provide the easiest way to
reach information resources that are available on the Web.The
number of Internet users has been growing dramatically all
over the world. Nevertheless, as for those who can not use
computers and the Internet, mobile search engine plays a great
role in offering the service. In this case, it quickly attracts
people's attention because it is unique. Mobile search is
extension to the search engine in the mobile terminal and is
also an important development trend in the future search
engine. The application in parallel computing technology is no
longer limited to scientific computing. This paper presents us a
novel mobile search engine architecture based on parallel
computing technology by combining parallel computing
technology with mobile search engine, so that it eases the
pressure from suddenly increasing system load and improves
the system's performance.
Keywords: search engines; parallel computing technology;
mobile search engine architecture; system load
I. INTRODUCTION
The traditional Internet search engines have become
indispensable tools for millions of users everyday and have a
strong base of users. Mobile device has the advantages of
access to information anytime and anywhere, so mobile
search is a hot issue in recent years[I]. Mobile search engine,
in this case quickly attracted people's attention because it is
unique. Mobile Search is also known as the wireless search,
which is an extension of search technology in the mobile
platform. It refers to a user through the mobile terminal,
using SMS, W AP, IVR and other access methods to access
to W AP sites, the content on the Internet, the content of
mobile value-added services and local information, etc.
Mobile search engines are based on the advanced mobile
communication technology in the mobile information
terminal to achieve the search engine system. Mobile search
combine mobile communications with search engines to
make mobile phone users who can not use computers and the
Internet make use of mobile phones and other mobile
terminals for mobile search[2].
For all mobile devices, the SMS service is a basic service.
Any mobile user can use SMS for search service. In recent
years, the researches on a SMS search have been the topic[3-
5]. But not like parallel computer technology-based mobile
search engine can serve for most mobile users.Parallel
computing technology originated from scientific computing,
but now its field in application is no longer limited to
scientific computing. a parallel computation is viewed as a
collection of interacting (data) parallel algorithms[6]. It has
been widely used in engineering calculations and many
emerging applications, such as data mining, stock trend
simulation analysis. Although business mobile search
engines can better respond to user queries, in a certain unit of
time when a large number of user requests the system, load
will be greatly increased, resulting in performance
degradation, or even work incorrectly.
In order to alleviate the above mentioned situation, this
paper presents a mobile search engine system based on
parallel computing technology and applies parallel
computing technology into mobile search engine, thereby
enhancing the mobile search engine capabilities.
II. SYSTEM MODEL
According to the system design goal, the system can be
divided into three logical parts.
I) data acquisition of the web search engine. This part
finds and searches pages information on the Internet by the
network search engine which is data sources of the whole
system.
2) data organization of the local cache database.This part
aims to cache and organize high-frequency query data and
build index library, so as to provide quick retrieval services
for theme query users.
3) data retrieval via interaction between SMS gateways
and mobile user.This part quickly retrieves documents
meeting the requirements in the index library by responding
to user's query request, and make relevance evaluation by
linking document with the query.The results will be sorted
out, and query results are delivered to the user.
Download full report
http://googleurl?sa=t&source=web&cd=2&ve...541250.pdf%3Farnumber%3D5541250&ei=SzsyTqiXKoyzrAfWxZDSCw&usg=AFQjCNGN6xMQVpKQvM0loRi5NNQfIzwKsQ&sig2=lRmDoRCKT-BE4kgd8oGK_Q
