I have read a paper which presents a high level description of an algorithm. Some steps of the algorithm are described in natural language and are hard to reproduce when trying to implement the code.
After reading other papers related to the original one I have managed to write all the subalgorithms in Pseudocode and implemented and tested the algorithm in C++.
In order to prevent others from going through the entire process again I would like to publish a scientific note which describes all the subalgorithms.
My work does not bring any novel contribution to the existing literature. It is just detailed description of an existing algorithm which might be useful to other researchers.
You could present and share at the International Congress
on Computational Engineering and Sciences. They just recently held a conference in Las Vegas but you can view the website at http://femtec2013.femhub.com/ Proceedings of FEMTEC 2013 appear in the Journal of Applied and Computational Mathematics (JCAM), an international journal with impact factor.
Its better to prepare as a short article or a note to the editor. Example: Information Sciences , IEEE Transactions on Cybernetics etc.
In case you actually provide a complete software, you could go for an applications note. (BMC Bioninformatics and OJ Bioinformatics) have something like that.
Nothing new, other than a new implementation is required.
If I can add another perspective, Ovidiu, why not create a data repository on your website with different problem instances and benchmarks? Presumably, you went through the implementation process because the problem which this algorithm solves is a) interesting, b) hard, or c) both. You can organize your library by size (does the algorithm scale well with problem size) and type of problem (is there something about certain instances that are harder to solve?) Using your implementation you could then create benchmarks and provide summary running times, etc. as benchmarks.
Hence, having a test library available for other researchers to validate *their* future algorithms is not only useful, but also likely to get you cited quite a bit!
- Badges
- Science topic
- Similar topics
- Papers