Neural Machine Unlearning for Privacy-Preserving Information Retrieval: Novel Methods and Corrective Mechanisms (Ref: CO/GC-SF5/2025)

This PhD project addresses the critical challenge of machine unlearning within neural information retrieval (NIR) systems. Contemporary NIR architectures require effective mechanisms to selectively remove specific data points while maintaining model performance – a fundamental requirement in privacy-conscious and adaptable information retrieval systems.

Background:

Neural information retrieval has revolutionised traditional search systems through deep learning implementations. While current approaches, including embedding-based architectures (DRMM, KNRM, DUET) and pre-training based frameworks (BERTdot, ColBERT), demonstrate impressive retrieval capabilities, they lack robust mechanisms for selective information removal. This limitation becomes particularly crucial when dealing with privacy requests, outdated information, or the need to correct erroneous data.

Machine unlearning in NIR presents unique challenges beyond traditional unlearning scenarios. The interconnected nature of query-document relationships in retrieval systems means that removing specific information can have cascading effects on model performance. Additionally, simply deleting information without providing suitable replacements can lead to noticeable gaps in search results, potentially triggering adverse effects such as increased attention to removed content.

Current unlearning approaches, primarily designed for classification tasks, prove inadequate for NIR systems due to the unique challenges of relevance scoring and ranking. Recent research indicates promising directions in developing specialised unlearning strategies for NIR, but fundamental challenges remain in balancing effective forgetting with maintained retrieval performance.

Research Objectives:

1. Develop novel machine unlearning methodologies specifically designed for NIR systems, focusing on selective forgetting while preserving model performance on retained information.
2. Create frameworks for evaluating unlearning effectiveness in NIR contexts, incorporating metrics for forgetting quality, retention performance, and ranking integrity.
3. Design efficient corrective mechanisms that enable the replacement of forgotten information with suitable alternatives, maintaining seamless user experience while ensuring privacy and accuracy.
4. Investigate the impact of different architectural choices and data characteristics on unlearning effectiveness, developing guidelines for unlearning-aware NIR system design.

This project advances the field of neural information retrieval by addressing crucial privacy and adaptability requirements, with direct applications in search engines, recommendation systems, and privacy-preserving information retrieval.

94% of Loughborough’s research impact is rated world-leading or internationally excellent. REF 2021

Start date

July 2025, October 2025, January 2026, April 2026

Supervisors

• Primary supervisor: Professor Georgina Cosma

Entry requirements

Applicants should have, or expect to achieve, at least a 2:1 Honours degree (or equivalent) in computer science or a related subject. A relevant Master’s degree and/or experience in one or more of the following will be an advantage: artificial intelligence, information sciences, mathematics with experience in programming.

English language requirements

Applicants must meet the minimum English language requirements. Further details are available on the International website.

Tuition fees for 2025-26 entry

• UK fee – To be confirmed Full-time degree per annum
• International fee – £28,600 Full-time degree per annum

How to apply

All applications should be made online. Under programme name, select Computer Science. Please quote the advertised reference number: CO/GC-SF5/2025 in your application.

To avoid delays in processing your application, please ensure that you submit a CV and the minimum supporting documents.

The following selection criteria will be used by academic schools to help them make a decision on your application. Please note that this criteria is used for both funded and self-funded projects. 

Please note, applications for this project are considered on an ongoing basis once submitted and the project may be withdrawn prior to the application deadline, if a suitable candidate is chosen for the project.

Apply now