Experimentation with the IEML artificial language for tagging articles in Stylo for information retrieval
The project aims to explore and experiment with the semantic possibilities offered by IEML, the artificial language created by Pierre Lévy, through the development of a program meant for the exploration of an ontology written in IEML for a corpus in SSH.
Issues
IEML is a language that allows a concept to be modeled semantically on the basis of characteristics that can be represented in matrix form. At the scale of an ontology, these explicit and semantic representations have the potential to reveal original and unexpected relationships between scientific articles. This is the purpose of the experiments carried out as part of this project using a wide range of keywords extracted from articles available on Érudit. By delivering semantic relations emerging from the exploration of an ontology, this project consists in giving back to users of scholarly literature research platforms an interpretable appropriation of new content against the logics of concentration of citation.
Technical challenges
The production of concepts in IEML is based on a complex identification of a broad semantic field. The first phase consists of identifying and specifying lists of words and concepts in IEML that can be useful for journals and setting up a selection process allowing a relatively simple handling. The second phase of the project focuses on the development of a user interface for concrete use cases rooted in the reality of the documentary research practices of researchers in the humanities and social sciences. A third and final phase will focus on deploying the program to allow its wide use coupled with search engines.
Research activities
- Study of the semantic fields of partner journals and identification of relevant and non-relevant keywords
- Creation of an IEML ontology
- Development of a prototype for the exploration of the ontology
- Scenario and alignment with Isidore's API
- API and/or pluggin deployment
Deliverables
Modular plugin or API that allows you to combine a search for documentary information with an unstructured but interpretable semantic exploration process.
