Call for Intership at IRIT

The objective of this internship is to develop/extend the system AGUTER ([1]) by making it interactive in order to allow users to iteratively model Scenarios with the system. Developing software still remains a very complex process consisting of different steps where several actors from the IT team and stakeholders are involved. Nevertheless, the testing step remains one of the biggest problems. We developed an approach to make the testing step easier, helping testers to design the user acceptance tests (UATs). This approach provides a good coverage of the situations that should be considered to test and elaborate a systematic and complete set of alternatives that need to be explored in order to test. Our approach is composed of two main steps: 1. Scenarios description. Scenarios ([2]) are stories about people and their activities to reach certain goals, starting from a setting and using some resources. Scenarios can be used to describe the application context (as a business case) and can also be used to describe software functionality (as use cases). Usually, scenarios are used to specify requirements and contain the user tests. In our approach, we propose to describe scenarios without user tests which will be generated automatically. Therefore, scenarios description (without test cases) are the inputs of our approach. 2. Test cases generation. The scenarios are transformed into an intermediate model (task/method [3]) so that they can be executed to produce user tests. User tests produced are presented in two forms: as an execution trees which contains all possible achievements of scenarios and as sentences in natural language.
Currently, the AGUTER system is composed of two main tools :
• a semantic wiki website for describing scenarios and to transform these scenarios in task/method model in a semi-automatic way;
• an Execution Engine which executes scenarios to generate user test cases (in the form of execution trees and sentences in natural language).

The objectives of this internship can be ranked in the following way by priority order:

1. Make an integrated system. The goal is to integrate the two tools (semantic wiki website and execution engine) into one system.
2. Develop a type of “software workbench” for scenarios modelling. Currently, the user inputs all scenarios and obtains test cases. We want to make this system interactive and iterative. For example, the user enters few scenarios, she/he can generate test cases, she/he can delete some test cases that are not interesting, and so on. An important functionality is to have an iterative process, that is the user inputs few scenarios, generate test cases, and then these test cases will automatically be integrated in the scenario descriptions (after eventually some deletions/modifications), after the process restarts. This design loop with the user is interesting because the test cases can give a new point of view on scenarios which can help to improve them.
3. Assist and analyze the scenarios design. This step aims at developing natural language, human-computer interface and graph theory tools to facilitate the design of scenarios.
4. Make a “collaborative software workbench” for scenarios modelling. This last objective is to extend the previous goals to group of users for a collective scenarios design.

Hosting team: ADRIA/SMAC at Computer Science Laboratory: IRIT - Toulouse - France
https://www.irit.fr/en/departement/dep-artificial-intelligence/adria-team/ https://www.irit.fr/en/departement/dep-interaction-collective-intelligence/smac-team/ Supervisors:
Pascale Zaraté : Pascale.Zarate@irit.fr
Guy Camilleri : Guy.Camilleri@irit.fr

Skills : Student in Master 2 at the University, an Engineering School or equivalent, in the field of computer science. Knowledge in natural language processing frameworks and semantics would be a plus. Good programming skills would be also highly appreciated. A strong capability to work in a team and communicate within an international team, both onsite and at a distance, would be very appreciated.

Starting date From April 2022.

Duration 5 months

Stipend Around 554 €/month

Place The internship will be carried out IRIT laboratory (Toulouse – France)

Contact Please send your CV and cover letter to all the supervisors.

References
[1] Antonelli, L., Camilleri, G., Torres, D. and Zarate, P. (2021), "AGUTER a platform for automated generation of user acceptance tests from requirements specifications", Kybernetes, Vol. ahead-ofprint No. ahead-of-print. https://doi.org/10.1108/K-04-2021-0252
[2] Alexander, I. and Maiden, N. (2004), “Scenarios, stories, and use cases: the modern basis for system development”, Computing Control Engineering Journal, Vol. 15 No. 5, pp. 24-29.
[3] Camilleri, G., Soubie, J.L. and Zalaket, J. (2003), “TMMT: tool supporting knowledge modelling”, 7th International Conference on Knowledge-Based Intelligent Information and Engineering Systems, KES 2003, Oxford UK, 03/09/03-05/09/03, Springerde la publication, pp. 45- 52, part I