International conference On Computer Science And technology

Gyancity Journal of Electronics and Computer Science
Vol.3, No.2, pp. 1-8, September 2018
ISSN:2446-2918 DOI: 10.21058/gjecs.2018.32001
1
Application of Page Ranking Algorithm Based on
Numbers of Link Visits in Web Recommendation
System for Online Business
Abstract- World Wide Web plays a vital role in global information service center. Online business is growing very rapidly by creating websites.
Website is made of number of webpages. Webpage is dynamic collection of hyperlink and usage information, providing rich sources of web data
for web mining. Due to exponential growth of dynamic information over the internet, information overload create big challenges for the researchers
in this area. One of the key components which ensure the acceptance of web page search service is the web page ranker a component which is said
to have been the main contributing factor of Google. This paper discusses the page ranking algorithms based on contents, structures and usages in
web mining, proposed page ranking algorithm based on number of links visit. Since every product has its own unique web page for its decryption
hence in selection of top most visited links can be applicable to create the recommendation list for the corresponding customer in online business.
Keywords-- PageRank, Web Mining, Recommendation System, Information retrieval, Web Graph, Online business.
1. Introduction
Web Mining is defined as the application of data mining techniques on the World Wide Web to find hidden information, This hidden
information i.e. knowledge could be contained in content of web pages or in link structure of WWW [8] or in web server logs. WWW
is a vast resource of hyperlinked and heterogeneous information including text, image, audio, video, and metadata. With the rapid growth
of information sources available on the WWW and growing needs of users, it is becoming difficult to manage the information on the
web and satisfy the user needs. Actually, we are drowning in data but starving for knowledge. Therefore, it has become increasingly
necessary for users to use some information retrieval techniques to find, extract, filter and order the desired information.
Search engine [2] receives users query, processes it, and searches into its index for relevant documents i.e. the documents that are likely
related to query and supposed to be interesting then, search engine ranks the documents found relevant and it shows them as results.
This process can be divided in the following tasks:
Crawler [11,12] is in charge of visiting as many pages and retrieves the information needed from them. The idea is that this information
is stored for the use by the search engine afterwards.
Indexing the information provided by a crawler has to be stored in order to be accessed by the search engine. As the user will be in front
of his computer waiting for the answer of the search engine, time response becomes an important issue. That is why this information is
indexed in order to decrease the time needed to look into it.
Searching: The web search engine represents the user interface needed to permit the user to query the information. It is the connection
between the user and the information repository.
Sorting/Ranking Due to the huge amount of information existing in the web, when a user sends a query about a general topic (e.g. java
course), there exist an incredible number of pages related to this query but only a small part of such amount of information will be really
interesting for the user. That is why the search engines incorporate ranking algorithms in order to sort the results.
Recommendation System There is individual page for every product and services provided by online business. Page ranking algorithm
rank every page which gives the popularity of the web page [9,10] that can be used to filter the most popular product.
Today is the era of online businesses. There are billions of small and large companies that provide the online facilities of sale and
purchase the product and services. Every product has the different links of its description. Hence PageRank is help to identify the
popularity of the product.
This paper is categorized into seven parts. First part discusses the idea about the web mining and its types. Second part is the related
study of page ranking algorithms; third part discusses the problems identification, forth part is implementation of page ranking algorithm
based on number of link visits, fifth part discusses its application in online business, six part is comparison analysis, and last part
discusses the conclusion of the paper.
2
Om Prakash Rishi
Director Research,
University of Kota, Kota,
Rajasthan, India
omprakashrishi@yahoo.com
1
Mahesh Kumar Singh
Research Scholar,
Department of Computer Science
and Informatics.
University of Kota, Kota,
Rajasthan, India.
maheshkrsg@gmail.com
3
Sumit Wadhwa
Assistant Professor,
Department of Information
technology & Management,
Greater Noida, U.P., India
sumitwadhwa1988@gmail.com

ISSN:2446-2918 DOI: 10.21058/gjecs.2018.32001
2
2. Web Mining
Extraction of interesting (non-trivial, implicit, previously unknown and potentially useful) information or patterns from large databases
is called Data Mining. Web Mining is the application of data mining techniques to discover and retrieve useful information and patterns
(knowledge) from the WWW documents and services web mining can be divided into three categories [1] as shown in figure 2.1.
a) Web Content Mining
b) Web Structure Mining
c) Web Usage Mining
2.1 Web Content Mining (WCM) describes the automatic search of information resources available online, and involves mining web
data content. It is emphasis on the content of the web page not its links. It can be applied on web pages itself or on the result pages
obtained from a search engine. WCM is differentiated from two different points of view: Information Retrieval (IR) View and Database
View. In IR view, most of the researches use bag of words, which is based on the statistics about single words in isolation, to represent
unstructured text. For the semi-structured data, all the works utilize the HTML structures insides the documents. For database view,
Web mining always tries to infer the structure of the Web site to transform a Web site to become a database.
2.2 Web Structure Mining (WSM) is used to generate structural summary about the Web sites and Web pages. The structure of a typical
Web graph consists of Web pages as nodes and hyperlinks as edges connecting two related pages. Technically, WCM mainly focuses
on the structure of inner-document, while WSM tries to discover the link structure of the hyperlinks at the inter-document level.
Web structure mining tries to discover the model underlying the link structures of the Web. The model is based on the topology of the
hyperlink with or without the link description. This model can be used to categorize the Web pages and is useful to generate information
such as similarity and relationships between Web sites. And the link structure of the Web contains important implied information, and
can help in filtering or ranking Web pages. In particular, a link from page A to page B can be considered a recommendation of page B
by the author of A. Some new algorithms have been proposed that exploit this link structure not only for keyword searching, but other
tasks like automatically building a Yahoo-like hierarchy or identifying communities on the Web. The qualitative performance of these
algorithms is generally better than the IR algorithms since they make use of more information than just the contents of the pages. While
it is indeed possible to influence the link structure of the Web locally, it is quite hard to do so at a global level. So link analysis algorithms
that work at a global level possess relatively robust defenses against spamming.
Fig 2.1 Taxonomy of Web Mining
2.3 Web Usage Mining (WUM) tries to discover user navigation patterns from web data and the useful information from the secondary
data derived from the interactions of the users while surfing on the Web. It focuses on the techniques that could predict user behavior
while the user interacts with Web. This type of web mining allows for the collection of Web access information for Web pages. This
usage data provides the paths leading to accessed Web pages. This information is often gathered automatically into access logs via the
Web server. CGI scripts offer other useful information such as referrer logs, user subscription information and survey logs. This category
is important to the overall use of data mining for companies and their internet/ intranet based applications and information access.
The three categories of web mining described above have its own application areas including site improvement, business intelligence,
Web personalization, site modification, usage characterization and page ranking etc. The search engines to find more important pages
generally use the page ranking. Implement PRNLV method use web structure and web uses mining technique to rank web pages and its
application in web recommendation system.
3. Related Work of Ranking Algorithms

ISSN:2446-2918 DOI: 10.21058/gjecs.2018.32001
3
The web is very large and diverse and many pages could be related to a given query. That is why a method/algorithm is used to sort the
entire pages subject to be interesting to a user’s query. All the algorithms consider the web pages as a directed graph in which pages are
denoted as nodes and links are denoted as edges.
3.1 PageRank Algorithm (PR): Surgey Brin and Larry Page developed a ranking algorithm used by Google, named
PageRank [2] after Larry Page (cofounder of Google search engine), that uses the link structure of the web to determine the
importance of web pages [7]. It takes back links into account and propagates the ranking through links. Thus, a page has a high
rank if the sum of the ranks of its back links is high. A simplified version of page rank is defined as follows
(3.1)
In the calculation of PageRank a factor c is used for normalization. Note that 0<c < 1 because there are pages without incoming links
and their weight is lost. Later PageRank was modified observing that not all users follow the direct links on WWW.
(3.2)
Where d is a dampening factor that is usually set to 0.85 (any value between 0 and 1), d can be thought of as the probability of users’
following the links and could regard (1 − d) as the page rank distribution from non-directly linked pages .Consider the following directed
graph [4]
Fig 3.1 Example graph
The PageRanks for pages A, B, C in figure 3.1 are calculated by using (3.2) with d=0.5, the page ranks of pages A, B and C
becomes:PR(A)=1.2, PR(B)=1.2, PR(C)=0.8
3.2 Weighted Page Rank Algorithm (WPR):
Wenpu Xing and Ali Ghorbani [5] proposed an extension to standard PageRank called Weighted PageRank (WPR). It rank pages
according to their importance not only consider link structure of web graph. This algorithm assigns larger rank values to more important
pages instead of dividing the rank value of a page evenly among its outgoing linked pages. Each outlink page gets a value proportional
to its popularity. The popularity is measured by its number of inlinks and outlinks.[6]
Where Iv,Ip and Ov ,Op represent the number of inlinks and outlinks of page v and page p respectively. The Page Ranks for pages A,
B, C are calculated by using (3.3) with d=0.5, the page ranks of pages A, B and C are PR (A)=0.65, PR (B)=0.93, PR(C)=0.60.
3.3Page Content Rank Algorithm (PCR):
Jaroslav Pokorny and Jozef Smizansky[3] gave a new ranking method of page relevance ranking employing WCM technique, called
Page Content Rank (PCR). This method combines a number of heuristics that seem to be important for analyzing the content of web
pages. The page importance is determined on the basis of the importance of terms, which the page contains. The importance of a term
is specified with respect to a given query q. PCR uses a neural network as its inner classification structure. The importance of a page P
in PCR is calculated as an aggregate value of the importance of all terms that P contains. For a promotion of the significant term and a
suppression of the others, the second moment is again used as an aggregate function [2]

ISSN:2446-2918 DOI: 10.21058/gjecs.2018.32001
4
Page_importance(P)=sec_moment({importance(t): t P}) (3.4)
3.4 Hyperlinked Induced Topic Search Algorithm (HITS) [13, 24]:
This algorithm assumes that for every query topic, there is a set of "authoritative" or "authority" pages/sites that are relevant and popular
focusing on the topic and there are "hub" pages/sites that contain useful links to relevant sites including links to many related authorities
as shown in figure 3.2.
Fig 3.2 Hubs and Authorities
Working of HITS: The HITS works in two phases Sampling and Iterative in the Sampling phase a set of relevant pages for the given
query are collected i.e. a sub-graph S of G is retrieved which is high in authority pages. The Iterative phase finds hubs and authorities
using the output of the sampling phase using following equations.
Where Hp is the hub weight, Ap is the Authority weight, I (p) and B(p) denotes the set of reference and referrer pages of page p.
4.0 Problems and Issues of the Page Ranking Algorithms
The main problems and issues of discussed page ranking algorithms are summarized as:
4.1 Rank quality of PageRank: The discussed ranking algorithms have shown a really high quality and the proof is that success of
Google (or they are still being used) successfully. However, some improvements can be done on it
4.2 Data Mining Technique of PageRank: PageRank algorithm used only Web Structure Mining and Web Content Mining technique; it
does not use Web Usage Mining, which may significantly improve the quality of rank of web pages according to users information
needs.
4.3 PageRank is Static in Nature: In PageRank algorithm, the importance or rank score of each page are static in nature. The rank changes
only with link structure of web.
5. Page Ranking Algorithm Based on Numbers of Link Visits (PRNLV)
PRNLV (Page Ranking based on Link Visits) based on Web Structure Mining and Usage Mining; it takes the user visits of pages/links
into account with the aim to determine the importance and relevance score of the web pages. To accomplish the complete task from
gathering the usage characterization till final rank determination many subtasks are performed such as
a)Storage of user’s access information (hits) on an outgoing link of a page in related server log files.
b)Fetching of pages and their access information by the targeted web crawler.
c)For each page link, computation of weights based on the probabilities of their being visited by the users.
d)Final rank computation of pages based on the weights of their incoming links.
e)Retrieval of ranked pages corresponding to user queries.
5.1 Calculation of Visits (hits) of links
If p is a page with outgoing-link set O(p) and each outgoing link is associated with a numerical integer indicating visitcount (VC), then
the weight of each outgoing link connecting to page p to page o is calculated by

ISSN:2446-2918 DOI: 10.21058/gjecs.2018.32001
5
where d is the damping factor as is used in PageRank, Weightlink is the weight of the link calculated by (5.1).The iteration method is
used for the calculation of page rank. Example fig 3.1 Taking d=0.5, these equations can easily be solved using iteration method the
final results obtained are:
PRNLV(A)= 1.08, PRNLV(B)= 1.26, PRNLV(C)= 0.66
Fig 5.2 Numbers of the Hits of the links
5.2 Experimental Evaluation:
To evaluate the performance and effectiveness of the proposed algorithm a prototype is designed a website is created using HTML,CSS
and Java script, wamp server is used as the server to store the activities of the users. The web graph of the web site as shown in figure
5.1. The data are collected by installing the model on the local host and uses are able to log the web site using their email as the ID,
visit to the different links. Every page has different URL with their visit count as shown in figure 5.2.Hence compute the page rank of
each link (page) using the formula (5.2) and arrange these links in its decreasing order of their page rank as shown in figure 5.3. Hence
the highest ranked page on the top of the list. The variation of page rank of the pages in the web site shown in the figure 5.4.
Fig 5.3 Page Rank Using (PR and PRNLV) Fig 5.4 Variation of PRNLV with PR
Variation of PRNLV with PR
Web Pages
PR
PRNLV

ISSN:2446-2918 DOI: 10.21058/gjecs.2018.32001
6
6. Application of PageRank in Recommendation System:
Every on line commercial company uses the specified pages for specified product that contains the specification of the product on their
web site. Every page has its own unique URL hence the popularity of the product can be computed by the popularity of the links that is
page rank of the page. Recommendation System tries to identify the user’s interest in the specific domain of contents based on their
previous experiences. When a user interact with the E-commercial site heshe offers a set of implicit or explicit information like clicks,
rating, comments etc. about his/her taste. Association Rule mining play a vital to associate the e-mail ID with the page with its rank that
can be helpful in creating the recommended list for the specified customer.
7. Comparisons of PRNLV with PR and WPR: From the table 1 it is clear that PRNLV is much better than PR and WPR.
Table 1 Comparison of Page ranking Algorithms
Algorithm
Parameter
PageRank (PR)
Weighted
Page Rank
(WPR)
Page Rank based on
Numbers Link-Visit
(PRNLV)
Description
Computes
scores at
indexing time.
Results are
sorted
according to
importance of
pages.
Computes
scores at
indexing time.
Results are
sorted
according to
importance of
pages.
Computes scores at
indexing time. Pages
are sorted according to
importance and
relevance.
Mining
Technique
Used
Web
Structure
Mining
Web
Structure
Mining
Web Structure
Mining, Web Usage
Mining
Rank Ranks are Ranks are Ranks are unequally
Distribution equally
distributed to
outgoing links.
equally
distributed to
outgoing links.
distributed among
outgoing links
according to their
probabilities of visit.
I/P
Parameters
Inbound links of
pages
Inbound links
and Outbound
links of pages
Inbound links,
Outbound links,
Visit Counts of
links.
Working
levels
n* n* N
Complexity O(log n) O(log n) > O(log n)
Nature of
Rank
Less dynamic
(rank changes
with link
structure )
Less dynamic
(rank changes
with link
structure )
More dynamic (rank
changes with visit
counts & structure of
links)
Relevancy of
pages
No No Yes
Importance of
pages
Yes Yes Yes
Quality of
result
Low High High

ISSN:2446-2918 DOI: 10.21058/gjecs.2018.32001
7
Advantages
Computation of
ranks with
minimum
effort and less
complexity.
Computation
of ranks with
minimum
effort and less
complexity.
Pages returned are of
high quality and
relevancy as user
feedbacks are taken
into account. Search
space can be very
much pruned as pages
are sorted according to
users’ information
needs.
Limitations
No relevancy of
pages is
considered in
rank
computation.
All links are
considered
equally
important.
No relevancy of
pages is
considered in
rank
computation.
All links are
considered
equally
important.
Extra effort on crawlers
to fetch the visit counts
of pages from web
servers. Extra
calculations to find the
weights of links.
8. Conclusion
Mining of knowledgeable data from a huge amount of data is very complex task, World Wide Web information play a vital role for
information collection and sharing. The ranking algorithms are used to search the relevance information in very efficient manner.
Different page ranking algorithms are used in different techniques. The PRNLV uses the user‘s browsing information in consideration
to calculate rank of a documents rather than link structure. Due to browsing information in consideration PRNLV system is more
dynamic than other ranking algorithms. The popularity of the web pages (products) computed by the page ranking algorithm can be used
for performance of the products in the on line business.
REFERENCES
[1] N. Duhan, A. K. Sharma, K. K. Bhatia, “Page Ranking Algorithms: A Survey”, 2009 IEEE International Advance Computing Conference
(IACC 2009) Patiala, India, 6-7 March 2009.
[2] A Broder. A taxonomy of web search. Technical report, IBM Research, 2002.
[3] Jaroslav Pokorny, Jozef Smizansky, Page Content Rank: An Approach to the Web Content Mining.
[4] A. Broder, R. Kumar, F. Maghoul, P. Raghavan, R. Stata. Graph structure in the web. In In Proceedings of the 9th International World Wide
Web Conference, 2000.
[5] W Xing,A. Ghorbani, “Weighted PageRank Algorithm”, Proceedings of the Second Annual Conference on Communication Networks and
Services Research (CNSR‘04), 2004 IEEE
[6] G. Salton . C. Buckley. “Term Weighting Approaches in Automatic Text Retrieval. In Information Processing and Management. Vol. 24, No.
5, pp. 513– 523 1998.
[7] S. Chakrabarti, B. E. Dom, S. R. Kumar, P. Raghavan, S. Rajagopalan, A. Tomkins, D. Gibson, and J. Kleinberg, Mining the Web‘s link
structure. Computer, 32(8):60–67, 1999.
[8] T. B.-Lee, R. Cailliau, A. Luotonen, H.F. Nielsen, A. Secret. “The World-Wide Web. Communications of the ACM”, 37(8):76–82, 1994.
[9] M.G Campana, F. Delmastro:” Recommernder Systems for online and mobile Social Networks: A survey”, Online Social Networks and
Media3-4(2017) , pp 75-97
[10] Q. Su , L. Chem:” A Method for discovering cluster of e-Commerce in terest patterns using Click- stream data”. Electronic
Commerce Research and Applications. 1567-4223 Elsevier-2014 http://dx.doi.org/10.1016/j.elerap.2014.10.002
[11] B. Pinkerton.” Finding what people want: Experiences with the web crawler”. In The second Internation WWW Conference Chicago, 1994.
[12] J. Cho, H. Garc´ýa-Molina, L. Page. “Efficient crawling through URL ordering”. Computer Networks and ISDN Systems, 30(1–7):161–172,
1998.
[13] C. Ding, X. He, P. Husbands, H. Zha, H. Simon. “Link analysis: Hubs and authorities on the world”, Technical report:47847, 2001.
[14] A. Borodin, G. O. Roberts, Jeffrey S. Rosenthal, P. Tsaparas. Finding authorities and hubs from link structures on the world wide web. In
World Wide Web, pages 415–429, 2001.
[15] S. L. Yong, M. Hagenbuchner Ah Chung Tsoi, “Ranking Web Pages using Machine Learning Approaches”, IEEE/WIC/ACM International
Conference on Web Intelligence and Intelligent Agent Technology, 2008.

ISSN:2446-2918 DOI: 10.21058/gjecs.2018.32001
8
Author’s Profile
Mahesh Kumar Singh received his B.Tech from UPTU Lucknow in 2004 , M.Tech. from Jamia Hamdard
University, New Delhi in 2012 and Perusing Ph D in Computer Science, Kota University on web mining and
its application to web business intelligence.
Dr. Om Prakash Rishi is Director of Research University of Kota and Head of Department of Computer
Science & Informatics. His Research areas are Artificial Intelligence, Intelligent Systems & CBR, Software
Systems and Agent technology, Cloud Computing, Information Security.
Sumit Wadhwa received his B.Tech from Kurukshetra University, Kurukshetra in 2009 , M.Tech. from
Kurukshetra University, Kurukshetra, in 2013 and research areas are Wireless Sensor Network & Artifical
intelligence.

International conference On Computer Science And technology

More Related Content

What's hot (15)

Similar to International conference On Computer Science And technology (20)

Recently uploaded (20)

International conference On Computer Science And technology