I discovered a mistake in my own published workshop paper, how do I handle errata?

Question

I am PhD student and the main author of a research-in-progress report in computer science. The paper is already published at a workshop (in the proceedings, this is not a journal paper) and was presented there a few months ago.

While continuing to work on the topic of the paper, I discovered a mistake I did in the already published workshop paper. A value that is referred to twice in the paper is wrong; it is still unusually high but lower than reported. The value is also mentioned in the summary of the paper. First I told my coauthors who sit in the next room in the same institute.

Now I want to publish an Erratum on my academic website.

My first idea is to publish a text file next to the download like that is labeled "erratum" (it is a single mistake). As an example for this "text file", I looked at the following webpage.

My second idea is to change the author version of my workshop paper that is downloadable as PDF file from my website. My current idea is to solve this via a footnote in the PDF version of the paper. So how would I handle that? Do I correct the value and add a footnote saying that the original published value was wrong and explain the mistake? The other option that I think is unsuitable is the following: to leave the wrong value in the paper and add a footnote that explains that this value was a mistake and give the correct value in the footnote?

Also, are there websites on how to handle errata for conference/workshop papers in technical fields like computer science? Please note also that in my field conference papers are quite important (e.g., the POPL conference) while in some other fields only journal papers count (sort of).

Answer 1

Mistakes happen. I would say that since it's a conference proceedings, therefore often the presented work is short of proofs, it's work in progress and it's very quickly baked, mistakes happen.

It is probably impossible to post errata to the paper. I would suggest the following steps:

• If you put it on arXiv, upload a new corrected version.
• If you put it on your personal webpage, add the errata there.
• Make sure that once you sumbit it for journal publication, the mistake is correted.

Answer 2

A published paper is a published paper so if your paper has a unique identifier (e.g. doi) and is considered an official publication, I do not think it is a good idea to make versions of the paper with differing information. That would just cause confusion since the reference would still be the same (you cannot publish the same paper twice with only a number changed).

If you publish with a regular journal, it would not be problematic to publish an errata in the same journal explaining the error. It is also possible to publish a correspondence where the error is discussed and hopefully convincingly shown not to alter the conclusions made in the original paper. Which way is applicable in each case can be discussed with the editor of the journal and also the best way to proceed. So contacting the journal and editor is a good first step to find out the best way forward.

You are of course entitled to provide this information on your own web site but the problem may be that it will be lost to most people reading your paper that way. Hence trying to get a published correction is the best way forward. A journal should be able to link the errata to the original article so that anyone interested in the paper will find the correction at the same time.

How this relates to a conference is unclear to me and is something you need to research for your specific case.