Abstract— In this paper the idea of mind map mining is
presented. We propose that information retrieved from mind
maps could improve academic search engines. The basic idea is
that from a mind map’s text, keywords can be retrieved to
describe research articles referenced by the mind map. So far,
we have not conducted any research on mind map mining.
Therefore this paper should only be seen as an early research
in progress paper, outlining the ideas and aiming to stimulate a
discussion. We start the discussion in this paper by presenting
some challenges that mind map mining is likely to face.

Index Terms— academic search engines, mind mapping,
mind maps, search engines, text mining
I. INTRODUCTION
Researchers often use academic search engines1 to search
for relevant work in their field. Usually, those search engines
allow the user to enter keywords and as result, all documents
are shown that contain the keywords. However, any
keyword-based search has to cope with various drawbacks
such as synonyms, homonyms and unclear or changing
nomenclature.
Different approaches exist to cope with these problems.
For instance, web search engines additionally index text from
linking websites (anchor text analysis): a search engine
shows website A for a certain keyword search even if this
keyword is not on the website, but if a linking website
contains this word in the link text. We propose to apply this
approach to mind maps and call it mind map mining.
Mind maps are a popular tool to structure and visualize
information. In the field of science they can be used for
drafting research papers. Some researchers do reference
scientific articles in their mind map by linking to
corresponding PDF files or BibTeX keys (see Figure 1 for an
example). Comparable to analyzing websites‟ anchor text,
the text in mind maps‟ nodes could be analyzed.
In the following section our idea is presented in more
detail. Then we describe some problems that seem likely to
occur. Finally, a summary and an outlook to future work are
provided.


Jöran Beel and Bela Gipp are with the Otto-von-Guericke University
Magdeburg, Department of Computer Science, ITI and SciPlore.org
(beel|gipp@sciplore.org).
Jan Olaf Stiller is with the FH Wolfenbüttel and SciPlore.org
(stiller@sciplore.org)

1 Sometimes it is distinguished between „academic search engine‟ and
„academic database‟ in the literature. However, these definitions are not clear.
Therefore only the term „academic search engine‟ is used in this paper for all
kind of services offering keyword based search for scholarly literature.
II. MIND MAPS AND THEIR POTENTIAL USE FOR KEYWORD
BASED SEARCH
Mind maps are diagrams which can be used to structure
and visualize information2. For the purpose of information
retrieval, a mind map does not differ significantly from a
scientific article. A mind map contains text and can reference
research articles, for instance, with BibTeX keys or by
linking corresponding PDF files (see Figure 2 for an
example). To draft a research paper in this way, special mind
mapping tools such as SciPlore MindMapping3 can be used
[5].


Figure 1: Extracting Keywords from a Mind Map

Analyzing anchor text should be applicable to mind maps.
If a node in a mind map references a document, the words of
the node (and parental nodes) could be assigned to that
document. Figure 1 illustrates this: The mind map contains
one node called “expert search” and child nodes link to
documents related to expert search (those with the red
arrows). However, many of these documents do not contain
the term „expert search‟ but other expressions. If search
engines would analyze mind maps and treat them as
„neighbored‟ documents, more relevant documents could be
found.
III. EXPECTED CHALLENGES
Analyzing references in mind maps is similar to analysing
references in scholarly literature. Accordingly, for mind map
mining, similar problems are to be expected as it is with
citation analysis. These problems are related to data
availability, robustness and timeliness, and are discussed in
the following sections.
A. Availability of Data
Citation analysis effectiveness is often limited due to a
lack of (correct) data [1, 2]: many research papers are not
cited at all; citation databases such as ISI Web of Knowledge
do not cover all available publications; and, due to technical

2 The ideas presented in this paper could be equally well applied to concept
maps and all other types of documents that link in some way to scientific articles
or other documents.
3 http://www.sciplore.org
Could Mind Maps Be Used To Improve
Academic Search Engines?
Jöran Beel, Bela Gipp, and Jan Olaf Stiller
Proceedings of the World Congress on Engineering and Computer Science 2009 Vol II
WCECS 2009, October 20-22, 2009, San Francisco, USA
ISBN:978-988-18210-2-7 WCECS 2009

difficulties, citations are not always recognized correctly
which leads to incorrect data in citation databases.
For mind map mining, availability of data is probably an
even more serious issue. To get a sufficient number of mind
maps, one could cooperate with mind map sharing websites
(e.g. http://www.mappio.com/) or with developers of mind
mapping software. For instance, we plan for a future version
of SciPlore MindMapping in which mind maps are directly
analyzed on the users‟ computers.
However, it is unclear as to the number of researchers who
actually use mind maps and how many are willing to share
their data. It seems likely that the number is rather low.
Nevertheless, mind mapping is a popular application. For
instance, the mind mapping tool FreeMind is downloaded
over a 150,000 times a month [6] and over 1.5 million people
use MindManager [7]. In addition, there exist hundreds of
books, blogs and websites about mind mapping. Therefore,
we are confident that sufficient data could be gathered.
B. Robustness of Data
Citations are often considered biased because authors
sometimes cite papers they should not cite and do not cite
papers they should cite [3, 4]. Accordingly, analyzing
keywords in citing articles might deliver irrelevant or at least,
not the most relevant papers.
The same seems likely to be true for mind map mining. If
researchers draft a paper with a mind map and include
references, these references probably are similar or even the
same as in the final paper. In addition, all social media
platforms must have to face spam and fraud as soon as they
become successful. There is no reason to assume this would
be different if mind maps were used for extracting keyword
for research papers. However, most social media platforms
find a way to cope with fraud and spam. If only mind maps of
„trusted‟ users were used, serious spam and fraud probably
could successfully be prevented. Trustworthiness of users
probably could be well determined in cooperation with social
networks, other community websites or by usage data of mind
mapping software.
C. Timeliness of Data
Publishing scientific articles is a slow process and it takes
months or even years before an article is published and
receives citations. Accordingly, extracting keywords from
citing articles can only be performed several months after
publication at the earliest.
Here, mind map mining seems to have an advantage. Mind
maps do not need to be published in journals or on
conferences. They could be analyzed the moment they are
created. This would enable the extraction of keywords
significantly faster than those with citation based
approaches.
IV. SUMMARY AND FUTURE RESEARCH
In this paper the idea of extracting keywords from mind
maps in order to improve academic search engines was
introduced: If a research paper is referenced by a mind map,
the mind map‟s text can be used to extract keywords
describing the referenced article. These keywords can be
used by academic search engines to improve the article‟s
classification.
We compared the proposed idea with classic citation
analysis and concluded that mind map mining is likely to be
inferior regarding data availability and maybe data
robustness but in contrast, superior in terms of timeliness.
Overall, mind map mining might prove to be a promising
field of research, having the chance to complement citation
analysis and text mining and enhance academic search
engines. However, there is a need for research since many
questions are unanswered:
How many researchers are using mind maps?
How many are willing to share them?
How exactly can keywords be extracted from mind
maps?
How should keywords extracted from mind maps be
weighted in comparison to keywords, for instance in
the document‟s title, abstract or full text?
Based on data that we will collect with SciPlore
MindMapping, we will conduct research to answer these
questions.

Figure 2: Mind map for a research paper (red arrows indicate links to PDF files; the tooltip displays a BibTeX key)
Proceedings of the World Congress on Engineering and Computer Science 2009 Vol II
WCECS 2009, October 20-22, 2009, San Francisco, USA
ISBN:978-988-18210-2-7 WCECS 2009

REFERENCES
[1] D. Lee, K. Jaewoo, M. Prasenjit, L. Giles, and O. Byung-Won. Are your
citations clean? Communications of the ACM, 50:33–38, 2007.
[2] M.H. MacRoberts and B. MacRoberts. Problems of Citation Analysis.
Scientometrics, 36:435–444, 1996.
[3] Jöran Beel and Bela Gipp. The Potential of Collaborative Document
Evaluation for Science. In George Buchanan, Masood Masoodian, and
Sally Jo Cunningham, editors, 11th International Conference on Digital
Asian Libraries (ICADL'08), volume 5362 of Lecture Notes in
Computer Science (LNCS), pages 375–378, Heidelberg (Germany),
December 2008. Springer. doi: 10.1007/978-3-540-89533-6.
[4] Jöran Beel and Bela Gipp. Collaborative Document Evaluation: An
Alternative Approach to Classic Peer Review. In Proceedings of the 5th
International Conference on Digital Libraries (ICDL'08), volume 31,
pages 410–413, August 2008.
[5] Jöran Beel, Bela Gipp, and Christoph Müller. „SciPlore MindMapping‟ –
A Tool for Creating Mind Maps Combined with PDF and Reference
Management. to appear.
[6] SourceForge. SourceForge.net: Project Statistics for FreeMind. Website,
2008. URL
http://sourceforge.net/project/stats/detail.php?group_id=7118&ugn=free
mind&type=prdownload&mode=year&year=2008&package_ihttp://sou
rceforge.net/project/stats/detail.php?group_id=7118&ugn=freemind&ty
pe=prdownload&mode=year&year=2008&package_id=0.
[7] MindJet. MindJet: About MindJet. Website, Juli 2009. URL
http://www.mindjet.com/about/.

Proceedings of the World Congress on Engineering and Computer Science 2009 Vol II
WCECS 2009, October 20-22, 2009, San Francisco, USA
ISBN:978-988-18210-2-7 WCECS 2009