Greetings fellow researchers,
Scientific progress is moving relatively fast these days, with plenty of research happening all over the globe, and a large body of research literature being continuously produced.
I have been thinking about ways that further accelerate scientific discovery and innovation, especially in the fields of Artificial Intelligence and Machine Learning. One aspect that can be improved is the time from inception of the idea for the research thesis to the final peer-reviewed well-documented paper. This time can be significant and the process can be very labor-intensive. Repositories like arXiv definitely address this issue by offering an intermediate stage for authors to share their preliminary results, and receive appropriate feedback on their work.
An additional step that we can take to further expand on the process is by introducing the Agile model that is recommended and employed in software development to scientific research. Scientific research mostly applies the waterfall method in developing scientific literature. The waterfall method is known for its rigidness and resistance to incorporating changes and new ideas. The idea here is that, instead of spending a very long time building something that might or might not work with a relatively small team, you deliver the product (the scientific paper) on multiple iterations, where each iteration is better researched than the one before. Doing this in a collaborative environment can produce high quality papers, while also allowing for swift adaptation to changes.
So, the proposal is: Just like in open-source software products, open the research topics to the public, and allow collaboration on the paper until it is mature enough for publication. All contributors to the paper will be appropriately and fairly accredited.
Advantages of this approach:
1-) More people collaborating on a topic can mean that more ideas and more labor can be put into it in a shorter amount of time, from writing the paper to running experiments, etc.
2-) It can allow for simple hypotheses to be rapidly tested, verified, and potentially evolve into theories and products faster than they would using the traditional methods.
3-) It can allow peer-reviewing in the very early stages as the paper is readily available at any time for feedback.
4-) It can set a low barrier of entry to people who are interested in research by allowing them to contribute to the literature, even if by small amounts.
5-) It records the history of the development of the paper, which gives a richer perspective on the paper.
6-) It allows for strong documentation of the paper and its software components (if any).
Disadvantages of this approach:
One potential disadvantage of using this approach is that: a large number of people collaborating on a project can create a messy and chaotic research environment. If you can think of more disadvantages, please share them.
How Open Source Research Hub works:
Any interested person with an idea paper, a hypothesis, or a project can post their project by emailing a draft of their hypothesis or proposal to opensrcresearch@gmail.com . Once approved, it will be posted to the GitHub account at https://github.com/opensourceresearchhub as a public repository, and the sender will be made the owner of the repository. From there, others who are interested can start collaborating on it.
Please leave any comments/suggestions/thoughts here, or directly using the email above.