A STATISTICAL APPROACH TO THE IMPACT OF FEATURED

ARTICLES IN WIKIPEDIA

Antonio J. Reinoso, Felipe Ortega, Jesus M. Gonzalez-Barahona

GSyC/Libresoft, Universidad Rey Juan Carlos, Madrid, Spain

Israel Herraiz

School of Engineering, Universidad Alfonso X el Sabio, Madrid, Spain

Keywords:

Wikipedia, Usage patterns, Trafﬁc characterization, Quantitative analysis.

Abstract:

This paper presents an empirical study on the impact of featured articles on the attention that Wikipedia’s

articles attract, and how this behavior differs in different editions of Wikipedia. The study is based on the

analysis of the log lines registered by the Wikimedia Foundation Squid servers after having sent the appropriate

content in response to the corresponding request submitted by any Wikipedia user. The analysis has been

conducted regarding the six most visited editions of the Wikipedia and has involved more than 4,100 million

log lines corresponding to the trafﬁc of September, October and November 2009. The methodology of work

has mainly consisted on the parsing of the requests sent by the users and on their subsequent ﬁltering according

to the study directives. Relevant information ﬁelds has been ﬁnally stored in a database for persistence and

further characterization. The main results of this paper are twofold: it shows how to use the the trafﬁc log to

extract information about the use of Wikipedia, which is a novel research approach without precedences in the

research community, and it analyzes whether the featured articles mechanism achieve to attract more attention

or not.

1 INTRODUCTION

Wikipedia stands as the most important wiki-based

platform and continues providing the overall society

with a great set of contents and media related to all

the branches of knowledge.

Most of the previous researching involving the

Wikipedia is concerned about the quality and reliabil-

ity of its contents (Priedhorsky et al., 2007; Olleros,

2008; Javanmardi et al., 2009) or focus on describing

its growth tendency and evolution (Tony and Riedl,

2009; Suh et al., 2009). By contrast, the study pre-

sented here focuses on the impact on the number of

visits of the consideration of a set of articles as fea-

tured. Moreover, the purportedly impact is analysed

for different editions of the Wikipedia.

The rest of the article is structured as follows: ﬁrst

of all, we describe the data sources considered and

then we introduce the methodology followed to con-

duct our work. After this, we present our results prior

to discuss them. Finally we propose some ideas for

further work and present our conclusions.

2 THE DATA SOURCES

This section aims to describe the data sources in-

volved in our analysis and particularly focuses on the

log lines containing information about the requests

submitted to the Wikipedia by its users. This kind of

information is offered by the Wikimedia Foundation

to universities or education centers interested in it for

research purposes. Therefore, the following sections

present the principal issues related to these log lines.

2.1 The Wikimedia Foundation Squid

Subsystem

Squid servers may be used to speed up web servers

by caching the contents repeatedly requested to them.

Under this approach, Squid servers are said to work

as reverse proxy servers.

As a part of their job, Squid systems do log in-

formation about every request they serve whether the

corresponding contents stems from their caches or, on

420

J. Reinoso A., Ortega F., M. Gonzalez-Barahona J. and Herraiz I..

A STATISTICAL APPROACH TO THE IMPACT OF FEATURED ARTICLES IN WIKIPEDIA.

DOI: 10.5220/0003101404200423

In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2010), pages 420-423

ISBN: 978-989-8425-29-4

 2010 SCITEPRESS (Science and Technology Publications, Lda.)

Table 1: Received ﬁelds from the Wikimedia Foundation

Squid log format.

Field Description

GMT time Current GMT

time

Request Total time spent

service to serve the logged

time (ms.) request

Request Request method

method (GET, POST, etc.)

URL URL containing

the request.

the contrary, are provided by the web servers.

2.2 The Wikimedia Foundation Squid

Logging Format

The Wikimedia Foundation Squid servers use a cus-

tomized format for generating their log lines. How-

ever, we do not receive all this information but just

those ﬁelds summarized in Table 1.

2.3 Featured Articles

Featured articles are considered the best articles all

over the Wikipedia. These articles have to meet a set

of criteria apart from the requirements demanded to

every Wikipedia article. These criteria cover from

a clear and comprehensive writing of the article to

a proper structure and organisation. Other aspects

such us stability, neutrality as well as length and ci-

tation robustness are also considered. However, the

purportedly quality of the Wikipedia featured article

has sometimes been subject of discussion (Lindsey,

2010).

In what our research is concerned, the consider-

ation of an article as featured can have a notable in-

ﬂuence over the number of visits to the article during

a period next to its promotion and may also affect to

the number of contributions received during the same

period.

3 METHODOLOGY OF THE

STUDY

The analysis presented here is based on a sample

of the Wikimedia Foundation Squid log lines corre-

sponding to three months: September, October and

November, 2009.

This analysis has focused just on the trafﬁc di-

rected to the Wikipedia project and to ensure that the

study involved its most mature and highly active lan-

guage editions, only the requests corresponding to the

most visited ones have been considered.

3.1 The WikiSquilter Project

The analysis of the received log lines consists on a

characterization based on a parsing process to extract

the relevant elements of information prior to a ﬁlter-

ing one according to the study aims and directives. As

a result of both processes, necessary data to conduct a

characterization are obtained and stored in a relational

database for further analysis.

Visits

Sept Oct Nov

1e+07 5e+07 2e+08

EN ES DE FR JA PL

Figure 1: Number of visits in the Wikipedia editions under

study.

Sept Oct Nov

Avg. num. of visits

0 200 400 600 800

Figure 2: Average number of visits for November’s featured

articles in the English Wikipedia.

Not all the information suitable to perform an

appropriate characterization can be obtained directly

from the log lines. More in the detail, the application

parser is devoted to determine the following informa-

tion elements:

1. The Wikimedia Foundation project.

2. The language edition of the project.

3. The namespace of the article.

4. The action requested by the user (if any).

5. The searched topic (if any).

6. The title of the article.

A STATISTICAL APPROACH TO THE IMPACT OF FEATURED ARTICLES IN WIKIPEDIA

421

Sept Oct Nov

0 1000 2500

English

Visits

Sept Oct Nov

0 1500 3000

Spanish

Visits

Sept Oct Nov

0 200 400

German

Visits

Sept Oct Nov

0 40 80

Polish

Visits

Figure 3: Boxplot of the visits to featured articles for the samples under study.

4 STATISTICAL ANALYSIS

In this section we present a statistical analysis of the

impact of the featured status on the attention that fea-

tured articles get.

4.1 Volume and Activity of the

Wikipedias

The visits to the considered editions of the Wikipedia

are shown in ﬁgure 1, and include all kind of visits,

like page views, editions, searches, etc.

4.2 Impact of Featured Articles on

Visits

Searching on the different editions of the Wikipedia,

we obtained a sample of all the featured articles cor-

responding to the month of October 2009. We ex-

tracted the number of visits for those articles dur-

ing the months of September, October and November,

with the aim of ﬁnding out what impact has the fea-

tured article mechanism on the visits that article gets.

We did all the analysis shown here using the GNU

R statistical package (R Development Core Team,

2009).

For instance, ﬁgure 2 shows the average number

of visits (or mean) for the November’s featured ar-

ticles of the English Wikipedia. At a ﬁrst glance,

it seems clear that featured articles attract attention

during the month where they are awarded the fea-

tured status, compared to the previous and following

months.

Figure 3 shows a boxplot of all the visits to the

featured articles of the samples under study. The box-

plot shows visits to articles that were awarded the fea-

tured status on October, and it tracks visits to those

articles the previous and following month (Septem-

ber and November). In the boxplots, the main box

shows the bulk of data (those values between the 25

and 75 percentile), and the median is highlighted with

a line inside the box. Outliers (values with very ex-

treme values) are marked with circles outside the box.

For instance, if we focus on the case of the English

Wikipedia, at a ﬁrst glance, it seems that level of vis-

its during October was higher than it was during the

months of September and November. However, in

September and November,the level of visits was sim-

ilar.

Interestingly, featured articles get more visits only

for the case of the English Wikipedia. For the rest of

samples, the differences in the median are very small.

Even more interestingly, for the case of the Polish

Wikipedia, featured articles get more visits the fol-

lowing month after being marked as featured. We do

KEOD 2010 - International Conference on Knowledge Engineering and Ontology Development

422

not explore this issue in this paper, although it clearly

deserves further and deeper research.

5 FURTHER WORK

The identiﬁcation and quantization of the requested

namespaces and the different types of actions may

serve to provide an exhaustive understanding of the

way in which the users are interacting with the

Wikipedia.

Finally, obtaining the searched topic when the re-

quest submitted by the user implies a research oper-

ation can help to analyse the use of the Wikipedia as

a searching engine or to evaluate the impact of the

searches on the most visited contents.

6 CONCLUSIONS

Our results indicate that we can observe that increased

attention only in the English Wikipedia, while in the

rest of the top six Wikipedias the featured status does

not affect the level of visits, neither in the month when

the status is awarded, nor in the previous or following

months.

We have found an exception to this, though. The

case of the Polish Wikipedia is interesting, because

featured articles gain attention in the month right af-

ter the featured status was awarded, this is to say, the

Polish Wikipedia’s response to the featured stimulus

is much slower than in the case of the English one.

REFERENCES

Javanmardi, S., Ganjisaffar, Y., Lopes, C., and Baldi, P.

(2009). User contribution and trust in wikipedia. In

Collaborative Computing: Networking, Applications

and Worksharing, 2009 CollaborateCom 2009. 5th In-

ternational Conference on, pages 1 –6.

Lindsey, D. (2010). Evaluating quality control of

Wikipedia’s feature articles. First Monday, 15(4).

Olleros, F. (2008). Learning to trust the crowd: Some

lessons from wikipedia. In e-Technologies, 2008 In-

ternational MCETECH Conference on, pages 212 –

216.

Priedhorsky, R., Chen, J., Lam, S. T. K., Panciera, K., Ter-

veen, L., and Riedl, J. (2007). Creating, destroying,

and restoring value in wikipedia. In GROUP ’07: Pro-

ceedings of the 2007 international ACM conference on

Supporting group work, pages 259–268, New York,

NY, USA. ACM.

R Development Core Team (2009). R: A Language and

Environment for Statistical Computing. R Foundation

for Statistical Computing, Vienna, Austria. ISBN 3-

900051-07-0.

Suh, B., Convertino, G., Chi, E. H., and Pirolli, P.

(2009). The singularity is not near: slowing growth

of wikipedia. In WikiSym ’09: Proceedings of the 5th

International Symposium on Wikis and Open Collab-

oration, pages 1–10, New York, NY, USA. ACM.

Tony, S. and Riedl, J. (2009). Is wikipedia growing a longer

tail? In GROUP ’09: Proceedings of the ACM 2009

international conference on Supporting group work,

pages 105–114, New York, NY, USA. ACM.

A STATISTICAL APPROACH TO THE IMPACT OF FEATURED ARTICLES IN WIKIPEDIA

423