Enriching the library subject headings with folksonomy
| Pages | 297-315 |
| Published date | 26 March 2020 |
| Date | 26 March 2020 |
| DOI | https://doi.org/10.1108/EL-07-2019-0156 |
| Author | Wei Yu,Junpeng Chen |
| Subject Matter | Information & knowledge management,Information & communications technology,Internet |
Enriching the library subject
headings with folksonomy
Wei Yu
Jiangsu Key Laboratory of Data Engineering and Knowledge Service,
Information Management School, Nanjing University, Nanjing, China, and
Junpeng Chen
College of Information Engineering,
Nanjing University of Finance and Economics, Nanjing, China
Abstract
Purpose –The purpose of this paper is to explore the potential of enriching the library subject headings
with folksonomyfor enhancing the visibility and usabilityof the library subject headings.
Design/methodology/approach –The WorldCat-milliondata set and SocialBM0311 are preprocessing
and over 210,000library catalog records and 124,482 non-repeatingtags were adopted to construct the matrix
to observe the semantic relation between library subjectheadings and folksonomy. The proposed system is
comparedwith the state-of-the-art methods and the parameters are fixed to obtain effectiveperformance.
Findings –The results demonstrate that by integrating different semantic relations from library subject
headings and folksonomy, the system’sperformance can be improved compared to the benchmark methods.
The evaluation results also show that the folksonomy can enrich library subject headings through the
semanticrelationship.
Originality/value –The proposed method simultaneousweighted matrix factorization can integrate the
semantic relationfrom the library subject headings and folksonomyinto one semantic space. The observation
of the semanticrelation between library subject headings andsocial tags from folksonomy can help enriching
the library subjectheadings and improving the visibility of the library subject headings.
Keywords Simultaneous weighted matrix factorization, Subject headings, Folksonomy, Social tags,
Library subject headings
Paper type Research paper
1. Introduction
Subject headings, the words or phrases selected from controlled vocabularies to describe the
subject or classification of document(s), are widely used in library catalogs and document
collections to facilitate information retrieval, indexing, classification and so on (Golub et al.,
2014). Beyond the original applications in libraries, subject headings have also been applied in a
variety of contexts for the enhancement of knowledge organization and discovery in the digital
environment (Vizine-Goetz et al.,2004). The subject headings directly contribute to effective
access and retrieval of tagged documents or Web resources (Beyene, 2017), as well as in
mapping and recommendation to other knowledge organization (O’Neill and Chan, 2003).
The authors thank the anonymous reviewers for their precious work. This work is supported by the
Ministry of Education of Humanities and Social Science Youth funded project (Grant No.
15YJC870002), the project of National Social Science Foundation of China (Grant No. 17ZDA291), the
project of China Scholarship Council (Grant No. 201906195008), the Project Funded by the Priority
Academic Program Development of Jiangsu Higher Education Institutions PAPD.
Library
subject
headings with
folksonomy
297
Received3 August 2019
Revised9 December 2019
Accepted5 February 2020
TheElectronic Library
Vol.38 No. 2, 2020
pp. 297-315
© Emerald Publishing Limited
0264-0473
DOI 10.1108/EL-07-2019-0156
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/0264-0473.htm
However, recent history witnessed a decline in the number of visitorsto library websites
and online catalogs (De Rosa, 2005). This in turn led to a reduction in the use of subject
headings by libraries. At the same time, the subject headings from controlled vocabularies
were also criticized for being slow to make relevant changes and too professional for
ordinary users (Lawson, 2009). It is reported that the comfortable use of some controlled
vocabularies requires proficiency and expertise, which might require users to have
extensive training and hands-on experiences(Lee and Schleyer, 2012). The reason might be
that the subject headings from the controlled vocabularies are specially developed by and
for librarians andother information specialists (Clements and Liew, 2016).
Since the mid-2000s, folksonomiesand social tagging have gainedpopularity in response
to increasinginteractivity and communityparticipation on the Web(Colombo-Mendoza et al.,
2018;da Silva et al.,2019). The term “folksonomy”is a portmanteau of “folk”and
“taxonomy.”Itrefers to the system in whichusers apply public tags to onlineitems, typically
to aid them in re-finding those items (Kipp et al.,2017). This practice is also known as
collaborativetagging or social tagging. Folksonomy is commo nly used in content repositori es
or social bookmarking projectsor websites, such as del.ici.ous,Flickr, CiteULike and so forth.
Supporters have stated that folksonomies can not only directly reflect the vocabulary of
users, but also reflect the population’s conceptualmodel of the information. Compared with
formal classification, tagging is simple and can provide much of the same value in
supporting browsing, discovery and many other applications. Some research has shown
that user tags can add value to the traditional subject headingsand enrich subject heading
categories (Garcia-Silva et al., 2012). Also, some projects allowed users to assign keywords
or tags to catalog records by themselves. For example, LibraryThing (Voorbij, 2012)isa
cataloging and social networking sitefor book lovers who can contribute tags assigned to a
book record which then appearas a tag cloud. Amazon.com also allowed users to tag books
with their own vocabularies.
However, this potential is seriously questioned because as a user-created vocabulary
opposed to a controlled one, folksonomy has faced criticisms on its free-form organization and
semantic limitations. Suffering from a lack of coherent categorization scheme and authority
control, folksonomy might lead to tag explosion or semantic ambiguity (Yi and Chan, 2009)and
hence offer a lower indexing quality (Garcia-Silva et al.,2012). Librarians are reluctant and
unfamiliar in assigning social tags to catalog records (Clements and Liew, 2016). According to a
study on LibraryThing tags for academic libraries, the large majority (80 per cent) are still
subject headings, and 9 per cent of the social tags are regarded as unconventional (Voorbij,
2012). In such a situation, the association and mapping of folksonomies to their semantically
corresponding subject headings could significantly enrich the vocabularies and improve the
quality of existing subject metadata, as well as alleviate the ambiguity and complexity caused
by uncontrolled folksonomy in the area of collaborative information systems. Furthermore, as
some research emphasized, implementing such integrations would improve the
interchangeability and augmentability of subject headings across platforms (Cantador et al.,
2011), and related Web resources are collocated for ease of discovery by the end user (Cantador
et al., 2008). Nevertheless, the manual mapping of users’vocabulary and the controlled one is a
laborious and fallible work. A study that looked at an experiment on medical students showed
that they needed additional courses to understand the gap between the formal medical subject
headings (MesH) and user tags (Lee and Schleyer, 2012).
To alleviate these issues, the objective of thisworkistodesignanddevelopanautomatic
method for mapping social tags to subject headings. Specifically, a novel simultaneous
weighted matrix factorization (SWMF) is proposed to integrate the subject headings from
OCLC catalog records and user tags from del.icio.us users’bookmarkingonalargescale.The
EL
38,2
298
To continue reading
Request your trial