Why in Japan would occur papers fraud. “LP-exam” software is in the sale of research fraud image.

Something topic of paper fraud problem now. Why is this such a problem would occur.

“<University of Tokyo> Papers scandal, to full-scale investigation investigation committee set up” from the Mainichi Shimbun on September 20, 2016 we wrote an article called. Even the University of Tokyo prestigious university of Japan’s leading discovered such problems, it has become a big topic as news.


What is this time the fraud research, and, why research fraud would be going on, about what there is and measures product, will continue to describe the perspective of author is a science graduate school graduate.

Research fraud is a blasphemy to prevent the development and reliability of science

From the Ministry of Education in 2014, guidelines on fraud research has been established. We will introduce to 3 couple of excerpts the basic idea of the guidelines.

① on which revealed the basic concept for the misconduct of the research activities, researchers to deter misconduct in research activities, while encouraging the efforts of the scientific community and research institutions, Ministry of Education, Culture, Sports, Science and Technology Ministry, distributing agencies and research institutions research in order to appropriately respond to misconduct by the person, in which each of the institutions indicate guidance on matters that should be maintenance.

② is fraud in scientific research, stacking the truth of the quest, are those contrary to the essence of science is a life to continue to create new knowledge, shake the confidence of the people of science, interfere with the development of science, blasphemy it is one which is one that can not forgive. Betrayal act for such science is meant to deny themselves the raison d’etre of the researchers, one in which to lose the trust of the scientific community.

③ implementation of scientific research are built on trust and mandate from society, if, on if or lost faded these trust and mandate, that will be foundation stand scientific research itself by collapses research involved who must all realize. Despite the severe financial situation, as prior investment in the future, also from the fact that promoting research and development by the Japanese Government in response to the public’s trust and mandate, ensure the fairness of the research activities is more demand even stronger. In addition, as a result of the scientific research of today it is performed by making full use of complex and diverse research methods or means to deepen the specialization as much as possible, although the scientific achievements and knowledge is going to dramatically increase, even the researchers to each other, from the fact that is a situation that is difficult to grasp the actual situation of research activities with each other, that the researchers advance the fair research has become more important than conventional.

<H26 from the Ministry of Education guidelines:


Among community development established science, are derived from the original national taxpayers, what kind of research has been conducted, such as what useful whether research has been conducted in a “trust” It has been promoted as a premise. Research fraud and may not be in Japan just a problem. Also it has become a problem all over the world. Please refer to this.

many research fraud is false and manipulation!


Figure 5. Percentage of papers containing inappropriate image duplications by year of publication.


Figure 7. Proportion of papers with image duplications by country.

URL:  http://biorxiv.org/content/biorxiv/early/2016/05/17/049452.full.pdf

Will the graph of the above is meant to do. Included in fact, it was to check the 2000 this paper for the past 20 years, even with a high impact factor (the higher the numerical value high-level Journal) magazine, also 3.8% percentage of paper that is used image with a problem It has been, yet also ranked Japan. According to the survey results shall be deemed to be replaced, 2 papers of the 10 papers that fraud is also reported that the paper coming out of Japan.

On the other hand, do you research the fraud there is any kind.

– Wrote the experimental results with different results

  • Have copy and paste the paper the results of other people

Graphs and data forgery, and falsification

– Image tampering

Etc., are several possible. Among actually research fraud is found cases, What kind of fraud is often.


Number of retractions since 1977 and cause of retraction

F.C. Fang, R.G. Steen, and A. Casadevall, “Misconduct accounts for the majority of retracted scientific  publications.” PNAS, October 1, 2012.

The top graph is a result of the analysis of this withdrawal thesis 2047 from 1977 to 2011, false (forgery, falsification, etc.) means that there were many by far. For what kind of fraud is often research in the field further, life science field of medical and biological systems has come out has been reported that most.

If you know the bad, it will be to become doing What is illegal.

That self-interest that put out the career advancement and good research work might be there as the cause, but no matter how excellent scholar’s, Will is forced to as the ability to make decisions become dull.

Research fraud is often the life science field. Image processing are common.

Earlier, it was described as illegal in many research in the life sciences in the research fraud. This is not perhaps as represented by the Obokata problem in STAP cells, handling a lot of image field, fraud in the field that must be subjected to the image processing in other words there are many.

The image fraud often present situation there is a background.

Image processing.

In recent years, hardware such as microscope and electron microscope, in particular the development of the camera is fiercely, its development does not have to mention that you have contributed very significantly in state-of-the-art science. However, even if caught is just how good image, even if taken just how fast the image is a large amount, it will eventually become the man of the scene to process the image.

And to process the image, for example, measuring the number of cells, and is typical, such as measuring the area, such as the analysis method is a status most of is self-taught is. We have it talk to a little research fraud, but the image analysis is the problem.

What is problem of the image analysis. For example, beginning in or would have much I researchers that may have read the research fraud guidelines of Education, Culture, Sports, Science and Technology Ministry has been established. Also, do you have much Japanese scientists who know the correct image processing method.

If you where I realized in college, probably students do not read only about one person If we had been 100 people. If you know the correct image processing method might be more less. For example, to process only a portion of the image, you can see even in elementary school, such as pasting a different image. However, not too excessively over in contrast adjustment, and say more precisely, I think in most of the researchers do not know, such as not to cause beautification.

Because now has progressed to the image processing software is more and more convenient direction associated with the microscope camera and it, simply if only was to clean the image not know the inside of the process, which has been pursuing a “non-scientific” part factors that cause will cause trouble will be increased to. Image processing of one-click, which was operated thought looked good, not a zero as the possibility that they become effective in the wrong image processing.

Where not good for Japan’s scientific community, it is that the education to be the root has been neglected. Since doing the premise and the research the Seizensetsu, if a problem occurs, will the chain to fraud resulting in rampant for measures only make the fire fighting is not done thoroughly.

Automatic image-manipulation detecting system “LP-exam series”

The University of Tokyo departure of venture companies, there is a strong company in the image processing of “El pixel Co., Ltd.”.

Therefore, in the image processing software, which is sold, there is a software called “LP-exam.”





And say how on earth this software, what is possible, to research image, where to whether there is a doubtful point in the image is detected automatically (some may not pick up depending on the image) research image for the fraud that has a function It is the software. Or on the Internet has come out anti-virus software, but something like a kind of security software in the field of research image.

As a product offline version, there is a cloud version and two, is a very innovative system called us to return the review results to automatically user to see the experts in the cloud version.

Offline version has no number limit, whether the image that is used to reference you want paper is correct, stand in very helpful when determining whether or not the image to be used in the paper that you will post is correct, in the heart of the relief material I think it is a great software that also. And also things that a function that can be batch processing for a large number image, this sounds pretty use likely.

Universities and research institutions, the private sector, I think in quite less unauthorized research on and have been equipped with these software in the standard so that it can be used by each person of the personal computer image. Once the research would happened fraud, trust and brand will fall to the ground in an instant. As part of the safety net, there should be a need to draw up such measures.

In order to protect the Japanese scientific community also, we pray also that such software is spread and introduced to the researchers of the country in order to protect the researchers.