| Description | A range of text analytic approaches can be applied to legal texts, and the results can be highly useful to citizens, law firms, and the government. Legal text analysis is a fast growing area of research, development, and commercialization.
In this project, you pick one or two text analytic approaches to apply to a corpus of legal texts (which can be provided, found, or created). The approach will be hybrid neuro-symbolic; that is, it is a blend of machine learning (e.g., Large Language Models) and rules (e.g., written in Prolog or another rule language).
One can chose from amongst:
* Search and information retrieval: storing and retrieving documents.
* Summarisation.
* Document clustering: grouping and categorising documents with respect to terms, phrases, paragraphs, or documents, using data clustering methods.
* Document classification: grouping and categorising phrases, paragraphs, or documents, using supervised data mining classification methods.
* Natural language processing: parsing and semantic representation.
* Information extraction: identifying, extracting, and structuring entities, concepts, and relationships from unstructured text into structured data (e.g., a knowledge graph).
* Sentiment.
* Deriving a decision tree from text which can be used for interactive input and determinations (inference). |
|---|