Yes! Even standard mathematical symbols and notations are captured in similarity index. The habit of using unconventional symbols and notations just to reduce similarity index is destroying the beauty and taste of mathematics.

The so-called similarity index essentially means "letter", "word", "symbol", and "sentence" repeating rate, not mathematical reasoning and not mathematical logic. Occasionally it is ridiculous that my full name and its abbreviations are also included in the similarity index!!! Therefore, we shouldn't believe in "machine" completely, we should believe in our eyes and brains.

According to your comment, it seems that even the biographical references would be harmful

Yes! Even standard mathematical symbols and notations are captured in similarity index. The habit of using unconventional symbols and notations just to reduce similarity index is destroying the beauty and taste of mathematics.

I gree with you Kwara Nantomah

A research paper must be evaluated by considering its content/contribution. The so called "similarity" forced us to change almost all symbols in a comic way as we are afraid of similarity. At least, Introduction not a part of similarity testing, so that the authors can include as much results which was previously discovered. Due to the plagiarism reason, we compel to short the introduction part.

Similarity must consider on main results, not in symbols and words.

You are right , I agree with you Dr. Nisar K S

but what we can do??????????

I agree that the beauty of mathematics is being destroyed by all these kinds of similarity checking software

The so-called similarity index essentially means "letter", "word", "symbol", and "sentence" repeating rate, not mathematical reasoning and not mathematical logic. Occasionally it is ridiculous that my full name and its abbreviations are also included in the similarity index!!! Therefore, we shouldn't believe "machine" completely, we should believe in our eyes and brains.

Agree! This program, when used blindly with non-professional persons, is really destroying. It considers even the names and institution of the authors.

I agree with views of Dr. Nisar K. S.

I agree with you Dr. Nisar K S

Similarity testing should not be applied to Introduction parts and symbols as Said by Dr. Nisar K. S. Most importantly, it should not be used blindly. One should not completely rely on software to decide plagiarism. After machine's job, it is the human who decides what is appropriate.

Thanks Prof. Feng Qi.

I agree with views of Dr. Nisar K. S.

You are certainly right.

What is the way forward then?

Should mathematician have a peculiar software for similarity index?

You are right.

I agree with views of Dr. Nisar K. S,

The journal's editors and reviewers should focus on the content and the quality of the research article.

I agree with the opinion of Dr. Nisar K. S.

I agree with you

I think Yes!

SINCE standard mathematical symbols and notations are captured in similarity index. The habit of using unconventional symbols and notations just to reduce similarity index is destroying the beauty and taste of mathematics.

Yes I do

It is not only for mathematics. It causes heavy confusion also for chemistry and chemical engineering where the same chemical equations could be used in in many publications. Actually, in this case it is very difficult to avoid, to escape, or to hide the similarities. For this reasons, this program should become more flexible to exclude mathematical and chemical equations from the similarity check.

agreed

I think yes. But how to solve this problem?

Some journals do have the criteria that the similarity index should be below this or that percentage. Sentences may be paraphrased in different words, however, the content and meaning may still remain the same. These softwares are not detecting plagiarism properly. I agree with Feng Qi that the mathematicians should rightfully follow and inherite traditional symbols, phrases, terminology, notions and notations from previous papers. The editors should check the contents and follow reviewing process to ensure that the results are correct and new.

Yes

One of the good things about iThenticate is that the software shows where the alleged plagiarism is. Then the editor or associate editor can make a decision on whether those individual instances constitute a natural mode of expression (such when describing how to introduce the stream function in 2D flows) or whether it is a sentence or two two taken from someone else's discussion. Thus a 40% score can be perfectly fine while a 10% score could be bad - it depends on the details, and the editors can estimate intent much better than an algorithm.

@Andrew S. Rees: Sir, we can not remove basic term (like potential flow, instability etc.) But Turnitin is showing that it is similarity with the published paper. What can be done to reduce similarity?

I do not believe that we need to reduce similarity.

The iThenticate score is a measure of similarity, but not everything that is computed to be similar is plagiarism. This is why a good editor plays an important role in deciding whether a 40% score is acceptable or unacceptable.

I believe that the only role that iThenticate plays is to identify those parts of a paper which need to be checked for plagiarism by the editor. The score is irrelevant, for as I mentioned, even a 10% score could correspond to blatant copying from elsewhere.

@Andrew S. Rees: Exactly Sir, we should not rely on only the similarity. But what I feel editors are rejecting the paper by saying that the similarity index is more than 10% in your paper. I am in favor that if someone copies the full sentence, it is plagiarism but the basic terms used in the paper do not fall under plagiarism category. Am I right?

I agree with you, Mukesh. The role for software such as iThenticate is merely to provide evidence of similarity. This isn't evidence of plagiarism, but it provides data from which plagiarism may be judged. If editors are rejecting papers because there is more than a 10% similarity, then those editors are acting incorrectly for they should be judging on the basis of plagiarism.

@Andrew S. Rees: Sir, last question; if we use mathematical symbol like rho, phi, mu etc. software is showing similarity, also if we recover the particular case of published papers, then again it is similarity. Is it plagiarism?

Similarity but not plagiarism. There are standard notations, as someone else has already pointed out, and the best clarity is attained by continuing to use such notations. This is not plagiarism. Definitely not. Plagiarism applies to the copying of someone else's *work* and making it look as if it is your own. Generally this is most likely to appear in Introductions and Discussion/Conclusions sections where whole sentences and even paragraphs can be taken from elsewhere.

By "recover the particular case of published papers" I presume that you mean that someone else's work is a subset of your own, and that the recovery of their work forms a check that your work is correct? If so, then that is fine although one needs to quote the paper whose work you are comparing with. It is essential for researchers to have objective checks that their work is correct - the use of multiple special cases, comparisons between analysis and numerics etc are all part of our proper duty to present work which has been checked carefully for accuracy.

I think that too many people have become obsessed with the concept of similarity and have thought that it is the same as plagiarism. We need to relax! If an editor confuses these concepts then they need to questioned.

Another thought...... Some say that the concept of self-plagiarism (where one uses one's own writing from another document without change) is self contradictory. Well, we can argue about that and I have seen strong views expressed on both sides of that argument. I am not going to be part of a debate on this, although copyright might be violated in such instances. However, if one has been writing papers for a long time it is quite natural to begin to develop one's own style. So, for example, the Governing Equations sections of, say, five different papers by the same author could look almost identical. That author might not have copied anything from paper to paper (although it is quite possible to do a copy and paste and then modify/rearrange slightly), but the similarity could have arisen simply because he/she writes the same way each time. No doubt iThenticate would give that a large similarity score, but that isn't plagiarism either.

Now I have work to do.....

If you will submit the paper in any Springer journal, you will face this kind of problem. Other journals are not like Springer. Suppose you are writting X follows some distributions name, then in Springer journal, this will come as similarity.

Basically it is a Nonsense software. I think all researchers should vehemently oppose this kind of non-sense software used some journals.

@Sanku Dey: there are various softwares like Turnitin, iThenticate, crosscheck etc. and they are being used to check similarity. But in these day editors are rejecting the paper on the basis of similarity (we know similarity is not Plagiarism). This is not a case of only Springer journals, everywhere these types of issues occurs.

Sanku Dey I disagree that it is a nonsense software. It provides useful information of similarity. It is then up to the editors to decide if those instances are plagiarism. If you have had a paper rejcted because of the similarity index being too high, then you should challenge the editor to say how much of that is plagiarism and you should be insistent that a proper assessment is made.

I would like to elaborate on what Professor D. Andrew S. Rees has mentioned: it is a matter of scale. The assessment of scholarly products is supposed to be conducted at an scale higher that what any code that determines similarities can conduct. That is due to the fact that even the most sophisticated machine learning algorithms still have significant difficulties in even distinguishing a picture of a cat from that of a table, leave alone the scientific forms. So, as long as the contribution is determined with respect to the human understanding of the phenomena and machine algorithms are less evolved than human brains, machines will be tools, not an authority. Now, what if the evolution of machines surpasses that of human beings? One might think of the idea as unrealistic. That is not true though: bare in mind that the evolution of machines is driven by the most powerful minds, not the average. As a result, it seems possible to have reviewing algorithms in the future that work better than a not so great reviewer. Such trend will make the understanding of machines, not human beings, the criterion. That fundamental change makes a new "space" as called in mathematics - a space that has already been evolving for a while.

The main problem is blindly trusting the outcome of these softwares and considering similarity and plagiarism the same. Most of the editors handle manuscript when it passes all checks, including the similarity check. Since the initial submissions are handled by editorial assistants, they return articles to the submitters because of insufficient knowledge of a particular scientific field.

Basic thing we are missing over here that almost in all cases Editorial Assistant looking the paper first and if he/she finds that similarity index is higher than their permissible limit (15%) they reject the paper. Secondly some Editors are also like that. In one of cases, paper is accepted by two reviewers but they found similarity index is higher than 15%. They sent me back for reducing. Subsequently, we have done the same and it was 14%. Even after that they rejected. I have written letter with turnitin similarity profile of my paper to the Editors two months before, till now they have not responded.

Thirdly, all references are coming under similarity index even in my case name and affiliation of my institute came under similarity. Forget about mathematical notations or universally accepted name etc...... I am facing every day... that is why in my point of view it is nonsense software...............

The awareness related to the difference between similarity and plagiarism should be emphasized .

As per now there is no software which can check plagiarism but plenty of software which can tell similarity index. In the view of above one can improve and represent his ideas in his own way and helps decrease the people who just provide information from copying from other website and pretend that to be them, this similarity index Will rule out them. This can never harm the Mathematics and institutions will also provide acceptance level of similarity and it checks for consecutive 3/5 words.

I will create Turnitin account and Grammarly premium account for plagiarism checking

Contact me by mail

[email protected]

Whats-app: +8801756240880

I will create Turnitin account and Grammarly premium account for plagiarism checking

Contact me by mail

[email protected]

Whats-app: +8801756240880

I think these programs are not qualified to deal with scientific papers in mathematics. To meet the requirements of these programs, most researchers resorted to making unnecessary modifications that resulted in distorting scientific articles. Therefore, these programs must be adjusted to suit this area. Until then, peer review could be relied on to identify scientific plagiarism.