Research Assistant – TANGO project
| Posting date: | 03 June 2025 |
|---|---|
| Hours: | Part time |
| Closing date: | 15 June 2025 |
| Location: | Swansea, Wales |
| Remote working: | On-site only |
| Company: | Swansea University |
| Job type: | Contract |
| Job reference: | SU00972 |
Summary
This is a fixed-term part-time role for 28 hours per week.
This role is a member of the Swansea University and PISA Human-Computer Interaction team on the TANGO project (https://tango-horizon.eu/). The TANGO project aims to develop the computational framework for effective hybrid decision-making making, based on:
A cognitive theory of mutual understanding and hybrid decision making.
Cognition-aware eXplainable AI (XAI) paradigms in support of synergistic human-machine interaction and collaboration.
A “Human-in-the-loop” co-evolution of human decision-making and machine learning models.
The impact on individuals and society will be evaluated on a pool of real-world use cases, namely supporting women during pregnancy, supporting surgical teams in intraoperative decision-making, supporting loan officers and applicants in credit lending decision processes.
You will contribute to two main tasks in the project:
Firstly, develop some small-scale demonstrators on top of the production TANGO API. These will be used to demonstrate the TANGO API functionality. This will cover particular activities related to the design and implementation of appropriate interfaces for TANGO-powered applications. We will put to the test coordinated decision-making processes involving virtual agents and humans, including assessing and contrasting options, suggesting courses of action, clarifying and establishing a decision-making process state, confirming or elaborating on communication, discussing problem-solving tactics, summarising courses of action, and identifying issues and alternatives.
Secondly, Developing Methods for Team Decision Making. We will employ a mixed-initiative approach to model and orchestrate agents and humans collaborating to facilitate effective decision-making. It will build on methods to facilitate decision-making partners in establishing common ground that is concrete and formal enough for agents but also allows nuanced human-focused argumentation where appropriate so that the shared decision richly accounts for different opinions and supporting arguments. The approach will identify the patterns of interaction between the user and agents including the co-evolution of users and agents during and between decision-making activities.
This role is a member of the Swansea University and PISA Human-Computer Interaction team on the TANGO project (https://tango-horizon.eu/). The TANGO project aims to develop the computational framework for effective hybrid decision-making making, based on:
A cognitive theory of mutual understanding and hybrid decision making.
Cognition-aware eXplainable AI (XAI) paradigms in support of synergistic human-machine interaction and collaboration.
A “Human-in-the-loop” co-evolution of human decision-making and machine learning models.
The impact on individuals and society will be evaluated on a pool of real-world use cases, namely supporting women during pregnancy, supporting surgical teams in intraoperative decision-making, supporting loan officers and applicants in credit lending decision processes.
You will contribute to two main tasks in the project:
Firstly, develop some small-scale demonstrators on top of the production TANGO API. These will be used to demonstrate the TANGO API functionality. This will cover particular activities related to the design and implementation of appropriate interfaces for TANGO-powered applications. We will put to the test coordinated decision-making processes involving virtual agents and humans, including assessing and contrasting options, suggesting courses of action, clarifying and establishing a decision-making process state, confirming or elaborating on communication, discussing problem-solving tactics, summarising courses of action, and identifying issues and alternatives.
Secondly, Developing Methods for Team Decision Making. We will employ a mixed-initiative approach to model and orchestrate agents and humans collaborating to facilitate effective decision-making. It will build on methods to facilitate decision-making partners in establishing common ground that is concrete and formal enough for agents but also allows nuanced human-focused argumentation where appropriate so that the shared decision richly accounts for different opinions and supporting arguments. The approach will identify the patterns of interaction between the user and agents including the co-evolution of users and agents during and between decision-making activities.