Student wanted: Implementing open-source quality rules

I presented two interesting resources in the last few months: Firstly, the INCOSE quality rules for requirements. The second is the open source tool Vale, which performs an automated quality check for requirements. These two resources could be perfectly combined to implement the INCOSE quality rules open source.

Why this makes sense, why it is interesting for SE Trends readers and what I would like to do with it is described below.

I am now looking for a student to implement this idea. This has the scope of a Bachelor's or Master's thesis.

Are you an employee in a company that could benefit from a quality check? An implementation with working students is also conceivable and I would be happy to accompany you if you are interested.

Are you in one of these situations or do you know people who might be interested? Then please pass on the information and get in touch with me.

The idea: Implementing open-source quality rules

Recently I introduced the INCOSE guidelines for writing requirements before. This guide is available free of charge (members) or for a small nominal fee in the INCOSE store. INCOSE not only allows itself a tool-based implementation of these rules, but even encourages tool manufacturers to do so.

Regardless of this, I had the open source tool Vale which can check the quality criteria of texts, including requirements. It therefore makes sense to use Vale to check the INCOSE quality rules.

Vale enables a very flexible configuration of the rules. Some of the INCOSE rules have already been implemented for Vale, for example the recognition of passive language. Some rules can be implemented quite easily. Some rules can probably not be implemented via a pure Vale configuration, but require programming.

Depending on the scope of work, this topic is suitable for a Bachelor's or Master's thesis. More on this below.

Why disclose results?

I would like to utilize the result in order to Semiant enable a simple quality check in any requirements management tool. This is just one of the many "skills" of Semiant. To achieve this goal, I would of course not have to disclose the results.

However, implementing open source quality rules has more advantages than disadvantages: Firstly, to put it bluntly, quality checkers are now a dime a dozen. The commercial value is therefore comparatively low. At the same time, there is the chance that a small community will form around this project, which would drastically increase the benefit of the result.

An open solution is also advantageous for students when looking for a job later on. Employers can not only view the work itself, the student can score points with a visible work on an open-source project.

Semiant: Far more than just a quality check

The vision behind Semiant goes far beyond a simple quality check, although that can certainly be a part of it. Semiant is a virtual quality assistant that has many configurable capabilities. But what they have in common is that we extract models from requirements texts that support specific tasks. I have listed a few ideas here:

For Semiant, an open-source implementation of INCOSE's quality rules would be a useful tool to demonstrate how Semiant works. The current plan is to offer the quality check free of charge.

Scope of the work

The aim of the work should be: "The analysis of the 41 INCOSE rules, as well as the implementation of a part of these rules with Vale". In concrete terms, this means as a minimum:

• Analysis of the 41 rules with regard to feasibility in Vale (fully, partially or not at all feasible).
• Implementation of the rules in Vale that can be implemented or partially implemented.
• Documentation on the use and configuration of the rules.

The above would be little more than a student research project. For a Bachelor's or Master's thesis, at least some of the following points would also have to be taken into account:

• Research of regulations for requirements quality, as well as the evaluation of INCOSE rules in the context of the regulations found.
• Researching software libraries for quality checks and evaluating the implementation of INCOSE rules.
• Extension of Vale (programming of plugins) to support the rules better than would be possible with a configuration alone.
• Study the effectiveness of the quality checker, ideally with an industry user

Nice to have:

• Implementation for German (in addition to English).
• Active community building of the results

Interested?

Then write to me or make an appointment for a short conversation in.

Photo by javier trueba on Unsplash