You are on page 1of 7

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/324482784

Internet Search Engine: Performance Evaluating the Google, Yahoo and Bing
Web Search Engine based on their Searching Capabilities

Conference Paper · February 2018

CITATION READS

1 2,304

2 authors:

Kamlesh Kumar Pandey Narendra Pradhan


Dr. HariSingh Gour Vishwavidyalaya (A Central University), Sagar, India 9 PUBLICATIONS 54 CITATIONS
37 PUBLICATIONS 122 CITATIONS
SEE PROFILE
SEE PROFILE

All content following this page was uploaded by Kamlesh Kumar Pandey on 12 April 2018.

The user has requested enhancement of the downloaded file.


Internet Search Engine: Performance Evaluating the Google, Yahoo and Bing Web
Search Engine based on their Searching Capabilities
Narendra Pradhan1 Kamlesh Kumar Pandey2

* Department of Computer and Information Science, Odisha State Open University, Sambalpur, PGDCS
Student KTNPUR,761003, Berhampur, India.
Mo: 9755390685, 8982182771, E-Mail: pradhan.narendra627@gmail.com.

** Department of Computer Science and Application , Dr. Harisingh Gour University, Sagar , Research
Scholar Amarkantak , 56789 Annupur, India
Mo: 7509341899, E-Mail: kamleshamk@gmail.com

Abstract² In Present day internet are most popular and into a subject classification, yellow pages scheme like a Arts,
used by everyone because all importance subject related Business, Computers and Internet, Entertainment, Government
information, data, news is given by WWW through internet. they are usually compiled by some type of logical order and
If any person can be wanted to any type of information in smaller database as a compared to indexes. Google and Bing
any time he searches to web search engine. Web Search are most popular Index web search engines and Yahoo is most
engine given the searched item related result in from list of popular directories web search engines.
link and person selected the one link at a time. Web Search A. Google Search Engine:-
engine given the some result link is related to search item Google is an American multinational technology
and some result link are not related for search item. In this corporation it given to Internet related services and include
research papers are cover to evaluating the performance of online advertise technologies, searching, android operating
top three web search engine Google, Yahoo and Bing on system, Mail, cloud computing, and software. Google is
their searching result. We evaluating the basic performance come to internet word in September 4, 1998.
factor precision and relative recall are on these search B. Yahoo Search Engine:-
engines. We divided on searching keyword in a simple one Yahoo is an American international technology
word, simple multi word and complex multi word group and business company and it come to January 1994 in internet
we taken on each group two searching keyword. Main aim word. Yahoo product and service is related to internet and
of this research paper is find out which web search engine is this service are a Yahoo News, Mail, Finance, Sports,
more relevant. Search, Messenger, Answer, Flickr, online mapping, video
sharing etc.
Keyword: - Web Search Engines, Precision, Relative C. Bing Search Engine:-
recall, Google, Yahoo, Bing. Bing is a product of Microsoft and launched on May
28, 2009 .Bing is also known as Live Search, Windows
Live Search and MSN Search. Product and service of Bing
I. INTRODUCTION
is Webmaster services, Mobile services, Developer
If any people can be wanting to any type of information at any services, searching, Toolbars, Gadgets etc.
time he goes to internet and searches this information through
web search engine in WWW. In present time many searching Performance of every web search engines has been
technique are available in web search engine and different type improving day by day with powerful search capabilities,
of web search engine used to own searching technique. Web technique of various types and the lack of restricted vocabulary
Search engines search to all information and all things in make it difficult for users to use web search engines
World Wide Web like a files, songs, videos, images, web sites, effectively. In this study we are attempting to assess the
weather information through various interfaces means through precision and relative recall of Google, Yahoo and Bing.
web search engine. In present day a lot of search engine are
available and each search engine are used a different language,
techniques, searching algorithm, web services and interface for II. PERFORMANCE EVALUATION OF GOOGLE, YAHOO
searching keyword. The National Information Standards AND BING WEB SEARCH ENGINE
society defines to boundary and access management is web
search engine. Performance evaluation of Google, Yahoo and Bing are
examined to during July 2015 to 20 October 2015 means we
Web Search engine is a web application program they run calculation the precision and relative recall for some selected
on the particular web address and this web address are called to search keyword between July 2015 to 20 October 2015. In this
website. User given a searching keyword on the specific web study, Google, Yahoo and Bing are given to own search results
search engine and web search engine are given the dynamically and this search result are categorized as five points. These
keyword related web page listing result on using on own points are first is more relevant, second is less relevant, third is
database. Web search engine are groups to categories first one LUUHOHYDQWIRXULVOLQNVDQGILYHSRLQWLVVLWHVFDQ¶WEHDFFHVVHG
is a Computer-generated indexes and second is a directories. on the basis of the following criteria and this criteria and points
Indexes base search engine are used to Spiders or robots search are given in the Chu & Rosenthal 1996[1], Leighton 1996[2],
programs and used large database they able to create a Ding & Marchionini 1996[3], Clarke & Willett 1997[4]. We
dynamically listing and Directories classify to web documents define new criteria on the bases on old criteria and calculation
NCDAMLS, 15-16 Feb 2018, Organized by Dept. CSIT, GGV, Bilaspur, C.G. ISBN-978-93-5291-457-9
65
on precision and relative recall of present time search. These a. Precision of Google
new criteria are identify the above five point and this criteria Total numbers of 3,75,00,00,000 sites are
are. founded for different six keyword and we selected to
x If the web page content is closely matched to subject 3000 sites for precision calculated. Following Table 1
matter of the search keyword then this type of web are shows the total number of more relevant web sites,
document is classify to more relevant and we given to less relevant web sites, irrelevant web sites, links and
2 number/score. sites cannot be accessed of Google in selection of
x If the web page content is not closely matched to 3000 sites. Clear for this table is 33.93% of sites are
subject topic but consists of some related aspects to the less relevant , 24.43 % irrelevant sites and 24.87% of
subject topic of the search keyword then this web page sites are more relevant. Precision mean of Google is
is classify to less relevant and we given to 1 0.907 found.
number/score.
x If the web page content is not related to the subject b. Precision of yahoo
topic of the search keyword then this web page is Total numbers of 739,700,000 sites are
classify as irrelevant and we given to 0 number/score. founded for six keyword and we selected to 3000 sites
x If the web page content is consisted of a complete for precision calculated. Following Table 2 are shows
series of links but some information is required then the performance of yahoo in above point more
this web document is classify as link and we given to relevant sites, less relevant sites, irrelevant sites, links
0.5 number/score. and sites cannot be accessed of yahoo in selection of
x ,IWKH ZHEVLWHRUZHEGRFXPHQWLVFDQ¶WEHDFFHVVHG 3000 sites. Clear for this table is 33.33% less relevant
or open for a particular URL then its web page is sites, 27.1 % irrelevant sites and only 24.63% of sites
FODVVLI\ WR VLWH FDQ¶W EH DFFHVVHG DQG ZH JLYHQ  are more relevant. Yahoo precision mean is 0.902.
number/score.
c. Precision of Bing
A. Precision calculation of Google, Yahoo and Bing Total numbers of 739,000,000 sites are
founded for different six keyword and we selected to
First factor of performance is precision so this section we 3000 sites for precision calculated. Following Table 3
calculation on precision of search engines for each of the are shows the performance of yahoo in above point
search keyword using this formula and used to five criteria (Eq. more relevant sites, less relevant sites, irrelevant sites,
1). links and sites cannot be accessed of Bing in selection
of 3000 sites. Clear for this table is 28.73 % less
Precision=Sum of the scores (number) of sites retrieved by a relevant sites, 21.47 % irrelevant sites and 31.1% of
Search Engine / Total number of sites retrieved sites are more relevant. Bing precision mean is 1.233.
(Eq. 1)

TABLE 1- Precision calculation of Google


Search Total number of Selec More Less Irrelevant links Sites Precision
keyword sites ted relevant relevant sites cannot
sites sites sites be
accessed
Simple one word queries
Computer 2,37,00,00,000 500 126 172 103 83 16 0.931
Multimedia 63,90,00,000 500 106 153 177 52 12 0.782
Simple multi word queries
Social 35,20,00,000 500 144 184 94 56 22 1.00
Marketing

Computer 29,30,00,000 500 156 186 72 63 23 1.059


science
Complex multi word queries
Evaluation of 5,05,00,000 500 101 154 183 56 6 0.768
Online
Marketing

Evaluation of 4,55,00,000 500 113 169 104 86 28 0.876


digital India
Total 3,75,00,00,000 3000 746 1018 733 396 107 0.907
(24.87%) (33.93%) (24.43%) (13.2%) (3.5%)

TABLE 2- Precision calculation of Yahoo


Search Total number Selected
More Less relevant Irrelevant links Sites precision
keyword of sites sites
relevant sites sites cannot be
sites accessed
Simple one word queries
Computer 527,000,000 500 132 183 96 72 17 0.966
Multimedia 62,800,000 500 98 195 142 56 9 0.838
Simple
NCDAMLS, 15-16 Feb 2018, Organized multi
by Dept. wordGGV,
CSIT, queries
Bilaspur, C.G. ISBN-978-93-5291-457-9
66
Social 102,000,000 500 123 112 237 19 9 0.735
Marketing
Computer 16,300,000 500 116 177 102 86 19 0.904
science
Complex multi word queries
Evaluation 21,300,000 500 121 169 123 73 14 0.895
of Online
Marketing

Evaluation 10,300,000 500 149 164 113 63 11 0.987


of digital
India
Total 739,700,000 3000 739 1000 813 369 79 0.888
(24.63%) (33.33%) (27.1%) (12.3%) (2.633%)

TABLE 3- Precision calculation of Bing


Search Total number of Selec More Less Irrelevant links Sites Precision
keyword sites ted relevant relevant sites cannot
sites sites sites be
accessed
Simple one word queries
Computer 5,370,00,000 500 106 151 141 89 13 0.815
Multimedia 539,00,000 500 96 118 192 82 12 0.702
Simple multi word queries
Social 308,00,000 500 201 99 119 72 9 1.074
Marketing

Computer 167,00,000 500 241 123 43 76 17 1.286


science
Complex multi word queries
Evaluation of 1,890,00,000 500 153 192 52 89 14 1.085
Online
Marketing

Evaluation of 116,00,000 500 136 179 97 72 16 0.974


digital India
Total 739,000,000 3000 933 862 644 480 81 0.989
(31.1%) (28.73%) (21.47%) (16%) (2.7%)

We selected the first top 500-result link given by Google, show in graph figure 2 in the base of searching Keyword and
Yahoo and Bing web search engine. We try to show precision. Finally Table 4 are summaries the total precision of
comparative precision analysis of Google, Yahoo and Bing simple one word, simple multi word and complex multi word
show in graph figure 1 in the bases of searching Keyword. group of Google, Yahoo and Bing and graph figure 3 are show
Comparative Performance analysis of Google, Yahoo and Bing to comparative precision on the basis this three group.

NCDAMLS, 15-16 Feb 2018, Organized by Dept. CSIT, GGV, Bilaspur, C.G. ISBN-978-93-5291-457-9
67
Figure 1- Comparative precision analysis of Google, Yahoo
and Bing
Figure:-3 Comparative precision analysis according to
word group

B. Relative Recall of Google, Yahoo and Bing


Second factor of performance is relative recall. Recall is
a retrieval system and it achieve all or most relevant
documents in the collection means recall is the ratio of the
amount of relevant records retrieve to the search engine and
total number of relevant records in the database.
Calculating on relative recall using this formula and this
formula (Eq. 2) given by Shafi & Rather 2005[5]

Relative Recall =Total number of web sites retrieve by a


search engine/ Sum of sites retrieved by the all search
Figure 2- Comparative Performance analysis of Google,
engine (Eq. 2)
Yahoo and Bing
a. Relative Recall of Google: - Total numbers of
3,75,00,00,000 sites are founded for different six
keyword. Google is given to relative recall is 0.72 in all
Table 4 ±Comparative precision of Google, Yahoo group but it given in Simple one word group have it
and Bing recall is 0.71, Simple multi word group have it recall is
Search Total Total Total Total 0.79 and Complex multi word have it recall is 0.29.
Engine number number number Precision b. Relative Recall of Yahoo:- Total numbers of
of of of 739,700,000 sites are founded for different six
Simple Simple Complex keyword. Yahoo is given to relative recall is 0.15 in all
one multi multi group but it given in Simple one word group have it
word word word recall is 0.14, Simple multi word group have it recall is
Google 0.81 0.99 0.82 0.90 0.14 and Complex multi word have it recall is 0.09.
Yahoo 0.90 0.82 0.94 0.89
Bing 0.76 1.18 1.02 0.99
c. Relative Recall of Bing:- Total numbers of 739,000,000
sites are founded for different six keyword. Yahoo is
given to relative recall is 0.14 in all group but it given
in Simple one word group have it recall is 0.14, Simple
multi word group have it recall is 0.05 and Complex
multi word have it recall is 0.61.

The relative recall of the Google, Yahoo and Bing is


calculated and show the Table 5 in the base of searching
keyword and graph figure 4 shows to comparative analysis.
We also try to summaries the total relative recall of simple one
word, simple multi word and complex multi word group of
Google, Yahoo and Bing in Table 6 and graph figure 5 are
show to comparative relative recall on the basis this three
group.

NCDAMLS, 15-16 Feb 2018, Organized by Dept. CSIT, GGV, Bilaspur, C.G. ISBN-978-93-5291-457-9
68
Table -5 Relative recall of the Google, Yahoo and Bing
Searching Google Yahoo Bing
Keyword Total No. of Relative Recall Total No. of Relative Recall Total No. of Relative Recall
Sites Sites Sites
Simple one word queries
Computer 2,37,00,00,000 0.69 527,000,000 0.15 5,370,00,000 0.16
Multimedia 63,90,00,000 0.85 62,800,000 0.08 539,00,000 0.07
Simple multi word queries
Social 35,20,00,000 0.72 102,000,000 0.21 308,00,000 0.06
Marketing

Computer 29,30,00,000 0.90 16,300,000 0.05 167,00,000 0.05


science
Complex multi word queries
Evaluation of 5,05,00,000 0.19 21,300,000 0.08 1,890,00,000 0.72
Online
Marketing

Evaluation of 4,55,00,000 0.67 10,300,000 0.15 116,00,000 0.17


digital India
Total 3,75,00,00,000 0.72 739,700,000 0.15 739,000,000 0.14

Table -6 Comparative Relative recall of Google, Yahoo and Bing


Search Engine Total number of Total number of Total number of Relative recall
Simple one word Simple multi word Complex multi word
Google 0.71 0.79 0.29 0.72
Yahoo 0.140 0.145 0.09 0.15
Bing 0.141 0.058 0.61 0.14

FIGURE:-4 COMPARATIVE RELATIVE RECALL ANALYSIS OF GOOGLE, YAHOO AND BING

FIGURE:-5 COMPARATIVE RELATIVE RECALL ANALYSIS ACCORDING TO WORD GROUP


all Bing precision is higher than Google and Yahoo, Google
III. CONCLUSION precision higher than Yahoo but we taken on group then
In this paper we presented the overview web search engine Yahoo is given to high precision 0.90 in Simple one word,
and Google, Yahoo and Bing search engine. Main aim of this Bing is given to high precision 1.18 in Simple multi word and
study is evaluating the performance and comparing to Google, Bing is given to high precision 1.02 in Complex multi word.
Yahoo and Bing web search engine on their performance. Second factor of performance is Relative Recall (most
Finally result of this study shows on precision of Google is relevant document) and this result is of Google is 0.72, Yahoo
0.90, Yahoo is 0.89 and Bing is 0.99 of all group. Then over is 0.15 and Bing is 0.14 of all group so Google have a higher
NCDAMLS, 15-16 Feb 2018, Organized by Dept. CSIT, GGV, Bilaspur, C.G. ISBN-978-93-5291-457-9
69
Relative Recall and Bing is given to lower Relative Recall in [9] .6ULQLYDV 3966ULQLYDV DQG $*RYDUGKDQ ³$ 6XUYH\
all group but we taken on group then Google is given to high RQ WKH ³3HUIRUPDQFH (YDOXDWLRQ RI 9DULRXV 0HWD 6HDUFK
Relative Recall 0.71 in Simple one word, Google is given to (QJLQHV´ ,-&6, 9ROXPH  ,VVXH  1R  0D\ 
high Relative Recall 0.79 in Simple multi word and Bing is Pages 359-364.
given to high Relative Recall 0.69 in Complex multi word [10] Felipe Bravo-0DUTXH] *DVWRQ /¶+XLOOLHU 6HEDVWLDQ $
group. At last Google is given to high relevant web article in R´ÕRV -XDQ ' 9HODVTXH] ³$ 7H[W 6LPLODULW\ 0HWD-
the bases of Relative recall and Bing is given to higher Search Engine Based on Document Fingerprints and
precision 6HDUFK 5HVXOWV 5HFRUGV´  ,(((:,&$&0
International Conferences on Web Intelligence and
Intelligent Agent Technology, pp-146-153
REFERENCES
[11] K.Srinivas, P.V.S. SrinLYDV$*RYDUGKDQ´:HE6HUYLFH
[1] &KX+ 5RVHQWKDO0  ³6HDUFKHQJLQHVIRUWKH $UFKLWHFWXUH IRU D 0HWD 6HDUFK (QJLQH´ (IJACSA)
World Wide Web: A Comparative study and evaluation International Journal of Advanced Computer Science and
PHWKRGRORJ\´ 3URFHHGLQJV RI WKH $6,6  $QQXDO Applications, Vol. 2, No. 10, 2011,pp-31-36
Conference, vol- 33, pp- 127-35.
[2] /HLJKWRQ +  ´3HUIRUPDQFH RI IRXU ::: LQGH[ [12] 'RUQ-DQG1D]7´6WUXFWXULQJ0HWD-search Research by
services, Lycos, Infoseek, Webcrawler and WWW 'HVLJQ 3DWWHUQV´ ,Qstitute of Information
:RUP´ 5HWULHYHG IURP Systems,Technical University Vienna, Austria;
http://www.winona.edu/library/webind.htm International Computer Science and Technology
[3] 'LQJ :  0DUFKLRQLQL *  ´$ &RPSDUDWLYH Conference; San Diego; April, 2008
study of the Web search service SHUIRUPDQFH´ [13] 6XEDUQD .XPDU 'DV´ 52/( 2) 0(7$ 6($5&+
Proceedings of the ASIS 1996 Annual Conference, vol ENGINES IN WEB- BASED INFORMATION
33, pp- 136-142 SYSTEM: FUNDAMENTALS AND CHALLEN*(6´
[4] &ODUNH6 :LOOHWW3  ´(VWLPDWLQJWKHUHFDOO 4th Convention PLANNER -2006, INFLIBNET Centre,
SHUIRUPDQFHRIVHDUFKHQJLQHV´ASLIB Proceedings, vol Ahmedabad, Mizoram Univ.,Aizawl, 09-10 November,
49 (7), pp- 184-189. 2006, pp-445-454
[5] Shafi, S. M., & Rather, 5$  ´3UHFLVLRQDQG [14] /X<0HQJ:6KX/<X&DQG/LX.´(YDOXDWLRQRI
recall of five search engines for retrieval of scholarly UHVXOW PHUJLQJ VWUDWHJLHV IRU PHWDVHDUFK HQJLQHV´ WK
LQIRUPDWLRQLQWKHILHOGRIELRWHFKQRORJ\´:HERORJ\ International Conference on Web Information Systems
(2), Retrieved from Engineering (WISE Conference); New York;2005.
http://www.webology.ir/2005/v2n2/a12.html [15] Yiyao Lu, Weiyi Meng, Liangcai Shu, Clement T. Yu,
[6] %7 6DPSDWK NXPDU DQG 603DYLWKUD ³(YDOXDWLQJ WKH and King-/XS /LX´ (YDOXDWLRQ RI UHVXOW PHUJLQJ
searching capabilities of search engine and metasearch strategies for
HQJLQHD FRPSDUDWLYH VWXG\´ DQQDOV RI OLEUDU\ DQG
[16] PHWDVHDUFK HQJLQHV´ ,Q :,6( YROXPH  RI /HFWXUH
information studies vol.57, june 2010, pp-87-97
Notes in Computer Science, pages 53±66. Springer,2005
[7] %LUDGDU % 6 DQG 6DPSDWK .XPDU % 7 ³,QWHUQHW VHDUFK
[17] 0HQJ : <X & DQG /LX . ³%XLOGLQJ HIILFLHQW DQG
engines:a comparative study and evaluation
HIIHFWLYH PHWDVHDUFK HQJLQHV´ ,Q $&0 &RPSXWLQJ
methodoORJ\´ 65(/6 -RXUQDO RI ,QIRUPDWLRQ
Surveys; 2002
Management, volume 43(3) ,2006 pp-231-241
[18] Javed A. Aslam and Mark Montague. Models for
[8] +RVVHLQ -DGLGROHVODP\´ INTRODUCTION TO
PHWDVHDUFK,Q6,*,5¶3URFHHGLQJVRIWKHWKDQQXDO
METASEARCH ENGINES AND RESULT MERGING
international ACM SIGIR conference on Research and
STRATEGIES: A SURVEY´ International Journal of
development in information retrieval, pages 276±284,
Advances in Engineering & Technology, Nov 2011,
New York, NY, USA, 2001. ACM.
ISSN: 2231-1963, pp-30-40

NCDAMLS, 15-16 Feb 2018, Organized by Dept. CSIT, GGV, Bilaspur, C.G. ISBN-978-93-5291-457-9
70
View publication stats

You might also like