Home

Welcome to the intermine-nlp wiki!

intermine-nlp is a library for generating InterMine PathQuery queries (JSON or XML) from natural language queries. There are many queries which are difficult or impossible to word in English, let alone translate into a PathQuery. For this project, it’s only worth targeting relatively simple query structures which users of the natural language interface are likely to care about.

Intermine is a database generation tool for genetic datasets. This library aims to provide a natural language front-end to Intermine, allowing users to interact with the database without needing to craft complex PathQuery queries.

InterMine databases are queried using a GraphQL-like query language called PathQuery, formatted in either JSON or XML. A web-based graphical tool for creating queries exists, but is still rather complex and unintuitive to use for those unfamiliar with the system. The maintainers of the InterMine project believe that an interface more similar to natural language would lower the barrier to entry for non-developers. Therefore, I’m proposing to work on adding a natural language to structured query translator, allowing users to more easily interact with their database.