top of page

Maya Ramanath

lastname at cse.iitd.ac.in

I am an associate professor in the Dept. of Computer Science and Engg. at IIT Delhi. I joined the institute in 2011 after a 5 year stint as a post-doctoral researcher in the Databases and Information Systems group at the Max-Planck Institute for Informatics in Saarbruecken, Germany.   I completed my Ph.D. from IISc, Bangalore.

 

​My recent research interests are in the areas of data management and information retrieval for multi-modal data. Previously, my group has worked on problems related to knowledge-graph construction, querying and applications for text and semi-structured data.

Courses

This semester (Dussera/Monsoon/Diwali) I am teaching COL764/7364: Information Retrieval and Web Search. We will start with the basics: from Boolean Retrieval to Language Models and then continue to more recent advances in the field, including Multi-modal and Multi-lingual retrieval in the era of LLMs.

​​

First class: July 28 (Monday), LHC 313.5, 2pm

​​

Look to Moodle for all announcements and course material.

Research

My research looks at improving the quality and efficiency of multi-modal processing and search. My group is specifically interested in:
  • Data management for multi-modal datasets. As part of our collaboration with students and researchers at Ashoka University, we have built a Datalake for multi-modal dataset management. The aim is to help researchers from diverse fields such as humanities, biology, economics, etc. do data-driven research by providing a platform to share high quality datasets. The Datalake encourages data science best-practices by automatically managing data provenance, versioning, visualisations, etc.
 
  • End-to-end processing pipelines (such as retrieval pipelines, OCR pipelines, etc.) that aim to process multi-modal corpora. We study metrics such as output quality, latency, energy efficiency and analyse how to increase the effectiveness of these pipelines.
​​
  • Apart from general-purpose processing pipelines, we also focus on specific application pipelines such as map processing/retrieval/question answering, processing of old, scanned archaeological reports with photos and layouts for retrieval. These are somewhat "niche" applications that are very helpful to researchers in areas such as archaeology, history, sociology, etc.

Previously, we have worked extensively on semantic web data management with a focus on knowledge-graph construction, querying and applications. We are hoping to bring many of the lessons learned to a new setting of multi-modal data management.

Students

Postdocs
Dr. Garima Gaur (2019-20, currently, researcher at INRIA)
 
​PhD
  • Nivedita Sinha (ongoing)​​
  • Prajna Upadhyay (graduated in 2021)
    • Thesis: "Construction and Applications of a Technical Knowledge Base")
    • First employment: Post-doctoral researcher at INRIA
    • Current employment: Assistant Professor, BITS-Pilani (Hyderabad Campus)
  • Madhulika Mohanty (graduated in 2021)
    • Thesis: "Techniques for Effective Search and Exploration over Knowledge Graphs")
    • First employment: Assistant Professor, IIT Dhanbad
    • Current employment: Researcher, INRIA-Saclay
  • Puneet Agarwal (graduated in 2019)
    • Thesis: "Discovering Relationships and Paths in Knowledge Graphs"
    • Current employment: Co-founder and CTO at RocketFrog.ai
  • Neha Sengupta (graduated in 2019)
    • Thesis: "Algorithms for Massive and Dynamic Graph Data Management"
    • Current employment: Principal Applied Scientist at G42
If you would like to work with me...

For IITD students: I take students for semester-long projects only (BTP or MTP). If you have completed courses in Machine Learning, Introduction to Database Systems, Information Retrieval and Web Search or a subset of these, send me an email.

For students looking for internships: I receive a lot of email from students asking for internships. It is not possible to respond to every email. I do offer internships during the summer break (minimum 12 weeks). If you are looking to do your project with me (typically lasting at least one semester), send me your CV. The requirements are the same as for IIT-D students -- courses in ML, DBMS and IR.
bottom of page