1. Thanthriwatta T.M

2.  The significance of effective and efficient
information retrieval.
 Usage of Internet and web technologies.
 Advent of WWW and web search engines.

3.  Generic web search engine cannot identify
the different needs of different customers.
 The search results should be personalized to
address this issue.
 Personalized Web Search concept was
introduced.

4.  The importance of the concept of
Personalized Web Search.
 Personal interest to study on this
technological area.

5.  Ability to identify the different needs of
different people who issue the same text
query for web search.
e.g. ‘Matrix’, ‘apple’
 Yahoo used this concept in 1998
 80 % of the users prefer to use personalized
web search engines.

6.  User profiling
 Client side implementation
 Server side implementation
 Content analysis
 Hyperlink analysis
 Community based PWS
 User location based PWS

7.  A separate user profile should be maintained for
each user.
 User profile consists with technical,
demographical and geographical informations
of users.
 Previously visited pages, total visit time, number
of visits , used links, age, gender, education, IP
addresses and bookmarks etc.

8.  Search engine has to maintain user profiles
by using its resources.
 Engine can use its all resources to optimize
the search results.
 Allocate a huge amount of memory and
computing processes to maintain millions of
user profiles.

9.  Users /clients are the responsible parties for maintaining
their user profiles.
 An installed software/plugin should be used to facilitate.
 Violation of privacy and security can be preserved as
much as possible.
 Cost of storage & computing processes are distributing
among users.
 Limitation of network bandwidth.

10.  This is under User profiling technique
 Check the similarity between web pages and
user profile details.
 User interested topics and title or content of
the web pages are much concerned.

11.  Most of the leading search engine use this
method
 Crawling and ranking concepts
 PageRank and Biased PageRank approaches

12.  Avoid the handling of separate user profile
for each user.
 Search engine has to find the users who have
similar kinds of interests.
 Efficient identification increases the
productivity of the collaborative web search.

13.  Groupization
 Give higher weights to pages that are relevant to
more members of the group.
 Hit-Highlighting
 Users’ keywords appeared within the title,
snippet or URL of each page are emphasized.

14.  Some users search things in a particular area. (by
implicit local intent queries)
 E.g. Italian restaurants in San Francisco
 It uses IP addresses mapping for identifying city
name, Zip code
 Issues
 Some names are ambiguous like Oakland
 Some have distinct meanings (Mountain view)

15.  Google Custom search engine
 Alpha search engine
 Google web history tool
 iGoogle & My Yahoo!
 Yoople! Collaborative web search engine

16.  This is an area, the concept can be used in a
practical manner.
 Identify the people in destroyed location by
analyzing user profiles.
 Alert people, before the disaster happens by
analyzing scientific stuffs.

17.  PWS is more time and effort consumable than
generic web search.
 Violation of privacy
 Ethical and security issues
 Users’ needs are not static.
 Handling more complex user profiles.

18.  How to overcome the limitations.
 Usage of location based PWS by covering
more geographical area.
 Query expansions
 Enhance PWS by using the commonsense
and folksonomy.

19.  The concept of PWS is an evolving study
area.
 Adding more techniques frequently.
 PWS can use for managing the real world
scenario like disasters etc.