Browse Prior Art Database

A system and method to recommend personalized learning/development needs based on automated code review findings/patterns Disclosure Number: IPCOM000247463D
Publication Date: 2016-Sep-08
Document File: 2 page(s) / 92K

Publishing Venue

The Prior Art Database


Background: On one hand, software developers go producing piles of lines of code that result in new features and good amount of it results in various types of bugs, in spite of all the sophisticated automated tools for source control and code reviews. On the other hand, we have LMS (Learning Management Systems), which enable an organization to plan, create & assign suitable learning plans to the employees (here we are talking specifically about software developers). Problem: One reason why the automated tools for code reviews/analysis are not effective, is that they review/analyze the entire code base or whatever 'files' have been presented to them, it predominantly goes by these 'file' concept (be it source code file or compiled binaries). Legacy code that carries a baggage of past issues, will be easy picks to ignore. Also centralizing this process (to rely less on individuals) gets defeated due to the same reason. We should be able to catch exactly those issues which are found in the recently changed code. And also allow us to draw patterns of teams/individuals' coding skills/practices.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 64% of the total text.

Page 01 of 2

A system and method to recommend personalized learning /development needs based on automated code review findings/patterns

LMS plans are predominantly gut/instinct based, not really directly linked/derived from individuals' skills. Above we have a potential source of discovering patterns/skill-gaps of software developers which need to be tapped into, to recommend suitable training/development plans.

Automated code review tools are integrated with source code version control tools in such a way that, the findings are restricted to 'lines of code' (rather files changes) in the selected date ranges/versions/developers. These findings will be used to create patterns of repeated mistakes being performed while coding per individual/team. And these historic patterns/trends will be used to create learning/development plans for individuals/teams, which will be fed into LMS (Learning Management Systems).

Proposed solution:
During the automated code review processes, we need to extract the findings of exactly the code that has changed
This again, should be picked in a selected time period
And of selected individual(s)

To accomplish the above, the automated code review tools need integration with version control systems (which exist in some way now)

Currently the integration between these automated code review tools and version control systems, again yielding the 'files' to be scanned

That'll again have the legacy issues left unattended previously o

Our proposed integration,...