Measuring book impact via content-level academic review mining

Pages138-154
DOIhttps://doi.org/10.1108/EL-08-2019-0184
Published date02 January 2020
Date02 January 2020
AuthorQingqing Zhou,Chengzhi Zhang
Subject MatterInformation & knowledge management,Information & communications technology,Internet
Measuring book impact via
content-level academic
review mining
Qingqing Zhou
Department of Network and New Media, Nanjing Normal University,
Nanjing, China and Department of Information Management,
Nanjing University of Science and Technology, Nanjing, China, and
Chengzhi Zhang
Department of Information Management,
Nanjing University of Science and Technology, Nanjing, China
Abstract
Purpose As for academic papers, the customarymethods for assessing the impact of books are based on
citations,which is straightforward but limited to the coverageof databases. Alternative metrics can be used to
avoid such limitations, such as blog citations and library holdings. However, content-level information is
generally ignored, thus overlooking usersintentions. Meanwhile, abundant academic reviews express
scholarsopinionson books, which can be used to assess booksimpact via ne-grainedreview mining. Hence,
this study aims to assess booksuse impacts by conducting content mining of academic reviews
automaticallyand thereby conrmed the usefulness of academicreviews to libraries and readers.
Design/methodology/approach Firstly, 61,933 academic reviews in Choice: Current Reviews for
Academic Libraries were collectedwith three metadata metrics. Then, review contents were mined to obtain
content metrics. Finally, to identify the reliability of academic reviews, Choice review metrics and other
assessmentmetrics for use impact were compared and analysed.
Findings The analysis results revealthat ne-grained mining of academic reviews can help users quickly
understand multi-dimensionalfeatures of books, judge or predict the impactsof mass books, so as to provide
referencesfor different types of users (e.g. libraries and publicreaders) in book selection.
Originality/value Book impact assessment via content miningcan provide more detail information for
massive users and cover shortcomingsof traditional methods. It provides a new perspective and method for
researches on use impact assessment. Moreover, this studys proposed method might also be a means by
which to measureother publications besides books.
Keywords Books, Impact assessment, Topic extraction, Review mining, Book feature extraction,
Library holdings, Sentiment identication
Paper type Research paper
1. Introduction
Citation analysis is commonlyused to assess book impact, but it depends on the availability
of citation indexes with adequate and appropriate coverage (Barilan, 2010;Krampen et al.,
2007). To overcome these obstacles to book impact assessment, several alternative metrics
are used to assess book quality,such as publisher prestige (Donovan and Butler, 2007), blog
citations (Shema et al.,2014)and library information (Ahmad et al.,2018;Torres-Salinas and
This work is supported by National Social Science Fund Project (No. 19CTQ031) and Youth Fund
Project of ISTIC (No. QN2019-11).
EL
38,1
138
Received5 August 2019
Revised9 October 2019
Accepted5 December 2019
TheElectronic Library
Vol.38 No. 1, 2020
pp. 138-154
© Emerald Publishing Limited
0264-0473
DOI 10.1108/EL-08-2019-0184
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/0264-0473.htm
Moed, 2009). However, none of these alternative metrics involves content analysis, and so
does not mine usersintentionsor motivations.
With the aim of more comprehensively assessing book impact,the role of book reviews is
currently under consideration (Kousha and Thelwall,2015, 2016). Book reviews include
academic reviews by scholars and online reviews by e-commerce or social media users.
Having previously speciedwhether and how book impact can be measured through online
book reviews (Zhou et al.,2016), this paper endeavours to mine academic reviews for book
impact assessment.
Unlike typical online reviews, academic reviews examine whether books provide new
knowledge to the eld, and how they relate to established theories (Nicolaisen, 2002). For
example, Dilevko et al. (2006) found that academic book reviews can help academic librarians to
stay current in their areas of specialization or to learn about unfamiliar areas. Zuccala et al.
(2014) used a machine-learning approach to code academic reviews as quality indicators for
book impact assessment. Kousha and Thelwall (2015) assessed whether academic reviews
couldbesystematicallyusedasindicatorsofscholarly impact, uptake or educational value for
scholarly books. However, different from research on online reviews (Zhou et al., 2016), prior
studies have coarsely analysed academic reviews, neglecting to mine ne-grained information
from review contents, and thus not comprehensively reecting a books impact. To ll this gap
in the literature, an assessment method was proposed for book use impacts based on ne-
grained mining on academic reviews of books. To prove the validity of the method, comparison
analysis was undertaken between metrics from academic reviews and existing metrics for
assessing use impact (i.e. library holdings). The analysis results should verify whether book
reviews can be used as auxiliary resources for library ordering and circulation decisions.
2. Related works
This paper focuses on using academic reviews to identify book impact. Hence, this section
examines two categories of related works: book impact assessment and review content analysis.
2.1 Book impact assessment
Citation is a traditional measurementfor evaluating the impact of academic books and prior
related studies are mainly based on citation databases. Thomson ReutersBook Citation
Index (BKCI) provides a potential new tool for bibliometrics (Leydesdorff and Felt, 2012).
Torres-Salinas etal. (2012) used the BKCI to analyse different impact indicators for scientic
books in social sciences and humanities from 2006 to 2011. Gorraiz et al. (2013) recognised
the BKCI as a rst step towards creating a necessary and reliable citation data source for
book impact assessments. Meanwhile, the citation facility of Google Scholar is commonly
used for research evaluation (Pitol and Groote, 2014). Kousha et al. (2011) found that online
citations in Google Books and Google Scholar can be sufciently numerous to support peer
review for research evaluation, according to signicant correlations among these two
metrics and Scopus. Abrizah and Thelwall (2014) contend that Google Books and Google
Scholar can be used for impact assessments of non-Western books. In addition, Scopus and
Web of Science are further important citation sources (Bakkalbasi et al., 2006). The above
analysis reveals that citation-based methods are commonly used, but depend on the
coverage of databases. Hence, citation-basedmethods are becoming inadequate for Web 2.0,
with researchers seekingmore online information to assess book impacts,such as libcitation
counts (White et al.,2009), library loans statistics (Cabezas-Clavijo et al.,2013) and topic
specicity (Daud et al.,2019). However, these metrics are numerical metrics, without
considering the relevant content informationof books. Therefore, it is difcult to reect the
purchasing or citing intentionsof users based on these metrics.
Measuring
book impact
139

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