Preserving diversity and inclusion is becoming a compelling need in both industry and academia. The ability to use appropriate forms of writing, speaking, and gestures is not widespread even in formal communications such as public calls, public announcements, official reports, and legal documents. The improper use of linguistic expressions can foment unacceptable forms of exclusion, stereotypes as well as forms of verbal violence against minorities, including women. Furthermore, existing machine translation tools are not designed to generate inclusive content. Therefore, the problem of spreading and promoting inclusive communication is challenging but also urgent.
INCLUSIVE COMMUNICATION CHALLENGES
- How to define non-inclusive communication?
- Whether Deep Learning techniques are suitable for modeling inclusive language?
- To what extent large-scale collections are suitable for learning pre-trained models?
- For adaptation to Inclusive Language tasks
E-MIMIC (Empowering Multilingual Inclusive comMunICation) is a joint effort of the research communities of linguistics and Deep Learning Natural Language Understanding in fighting against non-inclusive, prejudiced language forms. E-MIMIC aims at:
- Fostering inclusive communications in real-world scenarios
- Detecting and overcoming language inclusivity issues
- Grammatical asymmetry (silencing the feminine form)
- Semantic asymmetry (presence of stereotypes)
- Exploiting a deep learning pipeline to generalize and automate the process
- It leverages language models trained on purpose-specific corpora
- It can detect non-inclusive expressions and suggest inclusive alternatives
- Currently focusing on Academic and Public Administration Italian documents
- Can be adapted to different languages and types of documents
To the best of our knowledge, this project is the first attempt to empower inclusive formal communication with the goal of modeling users' meta-linguistic reflexive abilities.