Library usage mining in the context of alternative costs. The case of the Municipal Library of Prague

Pages558-576
Published date20 November 2017
DOIhttps://doi.org/10.1108/LHT-01-2017-0019
Date20 November 2017
AuthorPetr Hajek,Jan Stejskal
Subject MatterLibrary & information science,Librarianship/library management,Library technology,Information behaviour & retrieval,Information user studies,Metadata,Information & knowledge management,Information & communications technology,Internet
Library usage mining in the
context of alternative costs
The case of the Municipal Library of Prague
Petr Hajek and Jan Stejskal
Faculty of Economics and Administration, University of Pardubice,
Pardubice, Czech Republic
Abstract
Purpose The behaviour of a libr ary user is based on his a bility to evaluate the c onsumption of
library services based on al ternative market service s. The purpose of this paper is to study t he usage
behaviour in a public li brary building in the context of altern ative costs instead of a narrow focus o n book
circulation data.
Design/methodology/approach By library usage mining, using associative rules, the authors described
the behaviour of library users and identified the typical behaviour during the visits. The authors analyse the
results in the context of alternative costs assigned to the visits.
Findings The results confirm that some underused services, such as digital services, deliver
significantly great er benefit. The freque ncy of use, the duration o f visit and the number of it ems used
are associated with hig her alternative costs . There were no significa nt differences in alte rnative cost
within economic groups (e xcluding pensioners). This paper identified 41 fr equent patterns with diff erent
alternative costs.
Research limitations/implications The limitation of this study is the fact that data for library usage
mining were collected using a questionnaire survey.
Practical implications This may be particularl y important for both polic y makers and library
management. The first b eneficiaries are donors and patrons, wh o can learn about the benefit that librari es
bring to society. The proposed system for library usage mining also enables managers
to promote specific (effec tive) services, take steps to avo id readers leaving, and impro ve the adoption of
library services. It c an also be used to adapt the lo cation of library servi ces. Librarians, esp ecially those
who engage in acquisitio ns, may also use this inform ation in their evidence- based decisions about the
design of services.
Originality/value So far, there has been no relevant research on the economic aspects of extracted
behaviour patterns. Therefore, this study revealed userseconomic preferences using a questionnaire survey
that supplemented transaction data. The ability to describe usersbehaviour can provide library management
with enough information to realise evidence-based decision making.
Keywords Library services, Library users, User behaviour, Association rules, Alternative costs,
Usage mining
Paper type Research paper
Introduction
Library usage mining has become an increasingly important decision-support tool to
discover patterns of behaviour among library users in historical data (Nicholson, 2006).
Its greatest advantage is analysing user preferences and thus assisting librarians in
anticipating user needs. In addition, scenario analyses can be performed, and limited
resources can be effectively allocated to maximise customerssatisfaction (Siguenza-
Guzman et al., 2015).
Although a variety of methods have been applied for library mining, such as transaction
log analysis or prediction algorithms, association rules have been used predominantly,
because library usage mining is usually based on a large number of operational records.
These data are becoming increasingly rich and complex. The next issue to be addressed is
that discovering interesting user preferences using only statistical methods may be a
difficult task. As a result, complex statistical models must be created; on the other hand,
these are difficult to interpret. One solution to this problem is to employ an algorithm for
Library Hi Tech
Vol. 35 No. 4, 2017
pp. 558-576
© Emerald PublishingLimited
0737-8831
DOI 10.1108/LHT-01-2017-0019
Received 26 January 2017
Revised 24 April 2017
13 June 2017
Accepted 1 July 2017
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0737-8831.htm
558
LHT
35,4
extracting association rules. Thus, it is possible to discover easily interpreted rules covering
a large number of behaviour patterns in the underlying data.
However, far too little attention has been paid to the economic aspects of extracted
behaviour patterns. Indeed, to reveal userseconomic preferences is difficult and requires
using a questionnaire survey to supplement transaction data. To estimate the users
economic preferences, a sample of users is usually questioned to express preferences, and
thus both the explicit and implicit value of a public service can be reflected. This is also
known as a direct approach. Contingent valuation method (Mitchell and Carson, 1989) is
often used as a representative of direct methods. This approach works with a hypothetical
market, where the respondents express either their willingness to pay (WTP) the maximum
price or willingness to accept compensation for not being able to consume a service.
However, the results can be easily biased through non-meaningful budget restrictions,
inaccurate definitions of the evaluated service (information bias), non-acceptance of the
proposed scenario, etc. (Chung, 2008; Lee et al., 2010). To avoid the bias related to a
hypothetical market, we chose a method quantifying alternative costs. These can be defined
as costs library visitors must pay if the library does not provide services, and the visitors
have to replace them with market substitutes. For example, imagine a student who needs to
borrow a specific book for his university exam, but the library is under reconstruction.
Travel expenses to another open library and the value of the time spent are the alternative
costs. The expression of alternative cost is very easy for readers, it is easily measurable
(the reader has to pay it directly from his own pocket), and it is not distorted by subjective
evaluation (as is the case when using conventional contingent valuation method). If the
reader quantifies the alternative costs of selected library services, he is aware of the amount
of money and his answers in empirical survey would be provided more rationally.
In response to the aforementioned problems, in this study, a new hybrid library usage
mining system is proposed. In contrast to previous library usage mining systems, our
approach integrates the concept of customer value into the library usage mining process.
The academic contribution of this study can be summarised as follows: the novelty of this
study is to extend the conventional association rule problem by associating a customer
value with a transaction to reflect the users interest, enabling the discovery of valuable
patterns; another uniqueness of this study relates to the estimation of the customer value
using alternative costs, which is a novel approach to the evaluation of individual library
transactions. This methodology makes it possible to determine the benefits of the individual
library user behaviour patterns and thus contribute to the current knowledge in the domain
of library usage mining. The main purpose of this study is therefore to combine association
rules and alternative costs for mining library usage information to understand the economic
aspects of behaviour patterns. Tracing behaviour patterns and measuring their economic
value makes it possible to determine reader interest in certain services and whether these
services are used effectively. Taking the economic aspect of behaviour patterns into account
may therefore not only enhance our understanding of library user behaviour but also make
librariansdecision making more efficient.
The rest of the paper is structured as follows. The second section briefly reviews the
literature on library usage mining. The third section describes data and our research
methodology, including empirical survey characteristics. The fourth section presents the
results of our research. The final section concludes the paper and outlines several
managerial implications.
Library usage mining literature review
A recent comprehensive literature review on data mining methods applied to libraries by
Siguenza-Guzman et al. (2015) categorised related literature according to four main library
aspects, namely process/service analysis, quality of the collection and services, collection
559
Library usage
mining

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT