Extracting the semantics of places from natural language descriptions


3 Minute read

Human language (known as natural language or NL) poses many challenges for systems that use AI, for example, virtual assistants such as Siri and Alexa.

Conversing about places poses a specific range of challenges to current AI systems due to the flexibility and ambiguity of spatial language and the contextualised human knowledge of places. Context is critical for interpreting extractable terms such as place references and spatial prepositions, such as in locating something ‘at the back of the store’, or the notoriously difficult notions of distance, or relative sizes. But conversational patterns about places can be modelled or learned, translating them into formal interactive conversation patterns to support the development of intelligent assistance systems.

Project goals

The goals of this project are to:

  1. Investigate which spatial elements of natural language can be extracted together with relevant context, how these elements and their context can be represented in a database, and how these representations can be applied in querying the database
  2. Implement and test the methods required

Supervision team

*Click on the researcher's name above to learn more about their publication and grant successes.

Who we are looking for

We are seeking a PhD candidate with the following skills:

  • A masters in geoinformatics or computer science, experience with spatial data analysis or natural language processing
  • Excellent written and oral communication skills.
  • Demonstrated organisational skills, time management and ability to work to priorities.
  • Demonstrated problem-solving abilities.
  • The ability to work independently and as a member of a team.

Further details

The PhD candidate will profit from the combined expertise of the project supervisors, and the embedding into two research environments. Professor Stephan Winter and Professor Tim Baldwin at the University of Melbourne will contribute their expertise in the areas of spatial concepts in natural language, and Professor Sudeshna Sarkar will provide guidance in parsing language.

The candidate will be enrolled in the PhD program at the Department of Computer Science & Engineering at IITKg, and in the PhD program at the Department of Infrastructure Engineering at the University of Melbourne.

To apply for this joint PhD opportunity, and to view the entry requirements, visit How to apply.

First published on 16 November 2021.

Share this article