1 Introduction
Following the successful workshops at ECIR 20144 and 20155, respectively, this
workshop was the third in a series of events that brought together experts of
communities which often have been perceived as different ones: bibliometrics /
scientometrics / informetrics on the one hand and information retrieval on the
other. Our motivation as organizers of the workshop started from the
observation that main discourses in both fields are different, that communities are only
partly overlapping and from the belief that a knowledge transfer would be
profitable for both sides [
1
]. The first BIR workshop in 2014 set the research agenda
by introducing each group to the other, illustrating state-of-the-art methods,
reporting on current research problems, and brainstorming about common
interests. The second workshop in 2015 further elaborated these themes. This third
full-day BIR workshop6 at ECIR 2016 aimed to foster a common ground for the
incorporation of bibliometric-enhanced services into scholarly search engine
interfaces. In particular we addressed specific communities, as well as studies on
large, cross-domain collections like Mendeley and ResearchGate. This third BIR
workshop addressed explicitly both scholarly and industrial researchers.
2.1
Keynote
The keynote “Bibliometrics in online book discussions: Lessons for complex
search tasks” [
2
] was given by Marijn Koolen from the University of
Amsterdam. Koolen explores the potential relationships between book search
information needs and bibliometric analysis. The Social Book Search Lab is introduced,
which utilizes data from Amazon, LibraryThing (LT), the Library of Congress
and the British Library. LT discussions indicate some complex search tasks. Users
catalogue, tag, and relate books to each other. The hypothesis is that reviews,
catalogues, and discussion threads could be interpreted as (implicit) co-citation
and citation structures. Analyzing comments and reviews, several information
need patterns can be identified. Koolen also discusses how the data at hand can
be utilized for information retrieval.
2.2
Text and Reference Mining
In their paper “Weak links and strong meaning: The complex phenomenon of
negational citations” [
3
], Marc Bertin and Iana Atanassova designed a method
to extract negational citations from full-text publications. They revealed the
frequency distribution of such citations appearing throughout the regular IMRaD
structure of about 80,000 PLOS papers. Qualifying the polarity of citations has
many practical applications. This valuable knowledge might inform the scientific
community about papers attracting negative feedback that should be
reconsidered and potentially retracted.
In their paper “Towards a more fine grained analysis of sceintific authorship:
Predicting the number of authors using stylometric features” [
4
] Andi Rexha,
Stefan Klampfl, Mark Kröll, and Roman Kern aimed to chunk papers
according to stylometric features. The resulting segments were then attributed to the
corresponding author(s) listed in the byline of the paper (i.e., the individuals
who co-signed the paper). This contribution is likely to enhance paper/passage
retrieval by author name.
In their paper “The references of references: Enriching library catalogs via
domain-specific reference mining” [
5
], Giovanni Colavizza, Matteo Romanello,
and Frédéric Kaplan enhanced a digital library by collecting references from
domain-specific reference monographs in the Humanities. Their experiment on a
corpus dedicated to the history of Venice stresses the necessity of including such
overlooked references to improve search effectiveness in such corpora.
2.3
Tools for Bibliometric IR
In the paper “ Bibliometrics : a publication analysis tool” [
6
] by Rosa
PadrósCuxart, Clara Riera-Quintero, and Francesc March-Mir, the authors present a
bibliometric data management and consultation tool that can be utilized to
study and analyze an institution’s scientific activity. The tool is able to
generate bibliometric reports on scientific outputs at different analysis levels like
author, journal, and institution. The tool includes data from various sources
like WOS/Scopus and provides different indicators like productivity, visibility,
impact, and collaboration.
In the paper “Engineering a tool to detect automatically generated papers” [
7
]
by Nguyen Minh Tien and Cyril Labbé, the authors are focussing on detecting
fake academic papers that are automatically created. The authors work on
detection approaches based on distance/similarity measurement and introduce a
tool which is able to detect automatically generated papers, the SciDetect
system. The authors evaluate the SciDetect system against pattern matching and
Kullback-Leibler Divergence on three different text corpora.
2.4
IR Position Papers
In his article “Bag of works retrieval: TF*IDF weighting of co-cited works”,
Howard D. White proposes an alternative to the well-known bag of words model
called bag of works [
8
]. This model can in particular be used for finding similar
documents to a given seed one. In the proposed bag of works model, the tf and
idf measures are re-defined based on (co-)citation counts. The properties of the
retrieved documents are discussed and an example is provided.
In their article “On the need for and provision for an ‘IDEAL’ scientific
information retrieval test collection” [
9
], Birger Larsen and Christina Lioma argue
there is a need for test collections tailored to bibliometric IR. They discuss
several challenges coming along with creating such a collection (e.g., regarding
size, domain-specific dissemination and retrieval, realistic queries and relevance
judgements, pooling strategies as well as format). Furthermore, procedures to
create an ideal test collection are examined.
3
Outlook
With this continuing workshop series we have built up a sequence of explorations,
visions, results documented in scholarly discourse, and created a sustainable
bridge between bibliometrics and IR.
As a next iteration we will organize a Joint Workshop on
Bibliometricenhanced Information Retrieval and Natural Language Processing for Digital
Libraries (BIRNDL 2016)7 at the JCDL conference 2016. The BIRNDL
workshop will be co-organized together with the natural language processing group
of Min-Yen Kan, National University of Singapore, which includes a shared task
(the CL-SciSumm Shared Task8). The shared task tackles automatic paper
summarization in the Computational Linguistics (CL) domain.
7 http://wing.comp.nus.edu.sg/birndl-jcdl2016/
8 http://wing.comp.nus.edu.sg/cl-scisumm2016/