Post

Publication is an imperfect process

A publication often conveys the stereotype of a research project’s completion, or worse, acts as proof of truth and a definitive answer to a question. But it isn’t the case. A publication might not be a finished form of a work. Also, writing can always be improved. For most of us, it requires a significant amount of time and effort to put representations into words. Often, even when something is ready to be published, authors aren’t completely satisfied with their work. For many reasons, there can always be improvements.

Research, too, is an ongoing and imperfect process. It is not static; it’s organic and can sometimes be confusing. Additionally, it is often filled with misconceptions not only for students but also for researchers themselves. Rapid innovation and discoveries in the field of deep learning demonstrate that a technique may be considered part of the ‘state of the art’ one day, but the following day another technique may surpass its accuracy and performance scores. This doesn’t mean that the previous work isn’t worthy of consideration and study. The foundations of machine learning consist of various problem-solving approaches, and everyone benefits from learning about the history of a technique and the limitations that motivated the emergence of a new one.

When it comes to titles such as being graduated or being considered an expert in a field, I strongly oppose the idea that only certain individuals are permitted to publish their work. Take IT, where this notion is illogical compared to open-source software, where anyone can publish their code. These individuals publish their work with the intention of promoting improvements and even complete changes in implementations. Even if they don’t feel fully confident, they do it.

Most accomplished experts and professors in their field will admit there are always limitations, unexplored paths, and that they might be wrong. True publishing your work is undoubtedly uncomfortable; exposing you to comments and critiques. But once you become familiar with this process of exposure and learn not to take comments and recommendations as bad things, there are real benefits to sharing.

This post is licensed under CC BY 4.0 by the author.