The SHERPA project will develop and test the SELFIE Helper, an automated multilingual support system designed to facilitate schools in using the SELFIE platform and in on-boarding the overall SELFIE process.

The SELFIE Helper, designed to be fully and seamlessly embedded in the SELFIE platform, will take the form of a chatbot driven by Artificial Intelligence (AI) and Case-Based Reasoning. Ultimately, the SELFIE Helper is expected to significantly boost wide-scale adoption of SELFIE across Europe and at the same time reduce the human effort needed to scaffold constantly expanding SELFIE usage. These gains should contribute significantly to the overall sustainability of the SELFIE ecosystem.

Here is the draft architecture foreseen for the SELFIE Helper:

As you can see in the above draft architecture schema, the SELFIE Helper stores and manages data within its Knowledge Base. This contains a set of questions and answers organized into categories and topics: each category includes several questions and answers, while each question-answer pairing corresponds to exactly one category or topic. The Knowledge Base structure is illustrated in the Entity Relationship Model shown below.