[Monodevelop-devel] GSOC 2017 Projects

[Michael Viveros]

Hi,

I’m Michael Viveros and I’m interested in the Improve Auto-Documentation System project <http://www.mono-project.com/community/google-summer-of-code/projects/#improve-auto-documentation-system> for GSOC 2017 (mentor Mike Krüger).

Below are some questions.

1. What is an example of generating "documentation comments based on the member types and naming conventions"?
I read over the monodoc page <http://www.mono-project.com/docs/tools+libraries/tools/monodoc/generating-documentation/> about Generating Documentation but that was more about extracting documentation from code comments as opposed to actually generating documentation.

For example, maybe the auto-documentation system would take in something like `int addNumbers(int x, int y)` and produce “this function adds two integers”.
The output gets the “adds” part as a verb from the method stub and it gets the “two integers”part from the parameters to the method.

2. Where is the source code for the auto-documentation addin?

3. Do you have any initial ideas for how to use NLP to improve the auto-documentation system?
I know this is the main topic of the project but I just wanted to know if anyone has looked into it yet so I could have a starting point for my proposal. 
I saw on the Mono GSOC Trello board <https://trello.com/c/XncDuYUx/31-improve-auto-documentation-system> a link to Recurrent neural networks <http://karpathy.github.io/2015/05/21/rnn-effectiveness/> and it looked like they could be used to train a model to generate documentation after feeding it in sample methods.

Michael
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.dot.net/pipermail/monodevelop-devel-list/attachments/20170319/d48efd48/attachment.html>