The Excessive Power of Ctrl+C and Ctrl+V
in CS Research and Career Development
Various types of plagiarism are common in academia. Namely, (i) word-by-word or verbatim plagiarism: turning
contents into a carbon copy, (ii) paraphrasing plagiarism: copying the underlying meaning of the content in a
“smart” manner by slightly tweaking
the tone of sentences, (iii) idea plagiarism: capturing early/unpublished
ideas belonging to someone else, and
(iv) authorship plagiarism: masquerading as an original author of a different author’s work. Moreover, occasionally, reviewers may demand their own
papers to be cited (no matter whether
the references are relevant or not),
leading to a new type of plagiarism
that we call “citation plagiarism.” In
this article, we mention three real plagiarism stories, and show ways to reduce incidents of plagiarism. It is not
our intention to blame or shame any
one journal, editor, reviewer, university, student, professor, or author. Instead we will give an account of three
plagiarism stories to demonstrate
how insidious the problem is.
Recently, we were the victims of
plagiarism. We found clear evidence
of paraphrasing plagiarism in a published paper. It was a random online
search that caught our eye. A familiar
figure was identical to one published in
a paper from a couple of years ago. Curiosity lead us to discover 90 percent of
the content of that randomly searched
paper was paraphrased from not only
our paper, but others (available online)
too. We approached the editor-in-chief
of the journal of the newly published
paper, who found the same issue, leading to a retraction of the paper, after
The height of verbatim plagiarism
may be illustrated by the following
real story. Jure Leskovec, Anand Ra-jaraman, and Jeffrey D. Ullman wrote
a book entitled Mining of Massive
Datasets (2011). Since the book is
available online, someone by the
name of Seyed Hossein Ahmadpa-nah simply copied the entire thing.
Seyed only changed two elements of
the book: the title and the author’s
name. 1 Seyed used CreateSpace, Amazon’s book-production facility, to
produce the hard copies, and made
the fake book available on Amazon
Recently, we were assigned a paper to review for a well-regarded
journal. The writing and contents
of the paper were extremely poor,
which was not an unusual problem
until we noticed that the authors’
affiliation was listed as MIT. It was
unexpected to see MIT authors submitting a paper of that caliber, so
we investigated further. We discovered they were not actually enrolled
at MIT, or any other U.S. university.
Since we were curious to know what
went behind this blunder, we did not
reject the paper outright, but instead
asked for a major revision. The only
question that the authors responded
to was regarding their affiliation,
they claimed they mistakenly wrote
in the wrong affiliation.
After reading these stories, a natural
question arises in our mind: Why is plagiarism detection getting harder? There
might be various possible reasons:
1 See Prof. Jeffrey D. Ullman’s May 2017
Google+ post; https://plus.google.
1. Faulty tools. Many online plagiarism detectors can only find verbatim
plagiarism, not paraphrasing plagiarism. For example, Google Scholar does
not count a citation if a paper is not
properly cited in a new paper.
2. Fake reviewers. In some plots, authors first write a fake paper, and then
submit fake reviewers’ information
that turns out to assign the paper for review to the authors or to their friends. 2
3. A lot of papers and few reviewers.
Some top-notch journals/conferences
that promise to adhere to the standards
usually face a crisis of available reviewers. Most authors wish to submit their
papers to good journals; however, the
lack of expert reviewers significantly
impacts the review process.
4. Time and effort. Some students
wish to complete their degrees quickly
without devoting significant effort.
An identical notion is also applicable
to some faculty (usually at the beginning of their career), who wish to publish more to quickly earn a promotion.
These are the driving factors for plagiarism in terms of manipulating data, algorithms, or system settings.
5. Trust versus truth. Journals build
up the trust of researchers through impact factors, which are, occasionally,
false due to citation plagiarism and self-citation (the number of citations from a
journal article to articles published in
the same journal). Moreover, numerous
papers are submitted to journals for review. Hence, it is hard to accomplish a
uniform review process in any journal
with a good impact factor.
Transgender people are four times as
likely to have a household income
under $10,000, and twice as likely to be
unemployed as the average American.