“Wisdom of the crowds” and online information reliability. A case study of Israeli real estate websites
| Published date | 29 April 2014 |
| Pages | 417-435 |
| DOI | https://doi.org/10.1108/OIR-07-2013-0176 |
| Date | 29 April 2014 |
| Author | Maayan Zhitomirsky-Geffet,Yigal Maman |
| Subject Matter | Library & information science,Information behaviour & retrieval |
“Wisdom of the crowds”and
online information reliability
A case study of Israeli real estate websites
Maayan Zhitomirsky-Geffet and Yigal Maman
Department of Information Science, Bar-Ilan University, Ramat-Gan, Israel
Abstract
Purpose – The purpose of this paper is to determine whether the quality and reliability of websites’
content can be assessed through the lens of “wisdom of the crowds”. In particular as a case study the
authors examine the information supplied over time on several prominent Israeli real estate websites.
Design/methodology/approach – The Israeli real estate market was selected for the study, since
there are many large, popular and dynamic real estate websites that feature hundreds of thousands of
ads, representing most of the supply of real estate properties in the country. The authors built an
automatic, ontology-based system that downloaded advertisements from three selected websites
every two weeks for a number of months and checked for changes in these advertisements over time.
The authors conjecture that wisdom of the crowds is mostly reflected by the information changes on
the websites, since they indicate the anticipated market trends. Hence the autho rs developed a number
of statistical measures to comparatively analyse trends of information changes on these websites, and
assess their reliability compared to the actual market data and tendencies.
Findings – The primary results suggest similar information change trends amongst all the websites.
Surprisingly, although some properties did not sell over time, sellers generally did not lower their
asking price and were willing to wait. Sellers even raised their asking price, apparently in anticipation
of future price increases. Comparison of recurring trends among the websites with the trends of the
real market during the same time period and a few months after reveals that wisdom of the crowds is
only partially effective as an indicator and predictor of website content quality: it correctly reflects
the fluctuation in demand, but not in the prices.
Research limitations/implications – This study was conducted over a limited time period of five
months, and only in several cities in Israel. Additionally, since buyers are not explicitly represented
in these sites, their information behaviour was not analysed, although it undoubtedly influences
information changes performed by the sellers.
Practical implications – The practical contribution of this study is the ontology of the real estate
world. Its assimilation by real estate websites would promote the development of their sites and user
services. It would also enable ad sharing amongst the various websites and enable efficient searches
by search engines. In addition the tools and measures that the authors developed will allow continued
monitoring and analysis of user information change patterns.
Originality/value – Tothe best of the knowledge this is the first study to examine and compare real
estate websites’ quality and evaluate their information reliability as wisdom of the crowds.
Keywords Ontology, Online information reliability, Real estate websites, Website quality,
Wisdom of the crowds
Paper type Research p aper
Introduction
The internet provides a global information infrastructu re that enables the creation of
websites and applications for the purchase of goods and services, overcoming technical
problems such as geographic and time differences. Many e-commerce sites have become
extremelypopular (Amazon, eBay),and today e-commerceis a central componentof most
companies’ business activities.
Real estate is a market area that has successfully transferred to the virtual wo rld.
Todaymany coun tries have dozens of general real estate websites. In addition there are
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1468-4527.htm
Received 22 July 2013
First revision approved
27 August 2013
Online Information Review
Vol.38 No.3, 2014
pp. 417-435
rEmeraldGroup Publishing Limited
1468-4527
DOI 10.1108/OIR-07-2013-0176
417
“Wisdom of
the crowds”
local classified websites managed by private real estate agencies and sites that
specialise in specific types of properties, such as luxury apar tments. In Israel there are
close to 100,000 advertisements on the most popular classified sites, such as yad2.co.il,
at any given time. We can conclude that the online second-hand real estate market is a
popular model, changing dynamically and sufficiently large in scope (in relation to the
real market supply).
Unlike most online transactions, many real estate deals do not involve business
entities. Generally the transactions on these real estate websites are cond ucted between
private users (buyer and seller), although companies (such as real estate agencies and
construction companies) also present their listings on these websites. The website
services are generally free of charge and allow advertisement of real estate properties
for sale. In addition the service includes an interface that facilitates sophisticated
searches. Hence real estate websites’ content is assumed to reflect the “wisdom of
the crowds” viewpoint on the actual situation and trends of the actual market.
Consequently users search the ads on these websites not just to buy or rent properties,
but also to receive market updates and analyse the actual real estate market conditions.
Therefore the quality and the reliability of the content on the real estate websites
becomes quite crucial.
Numerous frameworks and me asures for online info rmation quality evaluation
have been presented in the past two decades (Eppler and Wittig, 2000; Klein, 2002;
Knight and Burn, 2005). The most prominent criteria from the literature are accuracy,
completeness, authority (known and visible author), relevance, maintainability,
accessibility, timeliness (being up-to-date) and usability. However, most of the measures
are not applicable to user generated content. Hence in a later review (Chai et al., 2009)
the authors highlight social content quality measures which are better fitted to
the assessment of content quality in the social web environment, e.g. internet forums,
blogs and social recommendation sites. The latter list of measures includes user
feedback, amount of data, reputation, objectivity, believability, consistency, value
added and security. However, quality evaluation of user generated commercial content
such as property advertisements on real estate websites were not addressed in these
studies.
Hence the goal of this work is to determine the credibility of the information
uploaded by thousands of users for other users on the e-commerce websites. To this
end we explored the possibility to regard the infor mation content of these sites as
wisdom of the crowds. This content quality can be then evaluated by the reliability of
the trends dictated by the collective intelligence on these sites during a period when
market experts had divided opinions about anticipated trends. We argue that in the
dynamic online environments, trends are represented by continuing repetitive patterns
of information changes over time. Hence to reveal these trends we explored the
information and the change patterns in information implemented by private sellers in
the online marketplace, focusing on the Israeli real estate websites as a case study.
In particularwe surveyed the actions, overtime, that advertisersimplemented to promote
the sale of their property, in order to identify permanent information change patterns.
Furthermore several quantitative measures were defined to analyse these changes and
assess the quality of the corresponding websites. The first aspect of the study was a
comparison of the various websites and identification of the differences between them
based on the suggested measures. In addition we analysed whether the website
informationwas influenced by the concurrent changes in the real market,and what their
impact was on the real market.
418
OIR
38,3
To continue reading
Request your trial