Gareth Jones's research interests are in the areas of natural language engineering including information retrieval and speech recognition, and intelligent multimedia systems. Dr. Jones's work integrates contributions from multiple disciplines to form novel information management technologies. Research areas in which he works include spoken document retrieval, scanned document retrieval, cross-language and multi-lingual information retrieval, context-aware retrieval, and video and image retrieval.

Dr. Jones is working within the Center for Digital Video Processing on systems for retrieval and management of video and multimedia content. In this work, he focuses on the indexing of video and audio data, how this information can be used to identify interesting content, and how users interact with multimedia content.

As well as developing systems which integrate multiple technologies, Dr. Jones is interested in more fundamental aspects of models for information retrieval and management. For example, he and one of his students have developed an effective new method for relevance feedback incorporating document summarisation, and ongoing work is exploring the application of natural language processing to enhance information retrieval models.

Dr. Jones's research in multimedia is currently centred on a project with a Ph.D. student looking at the automatic characterisation of non-linguistic audio features for characterisation of video data. This work is seeking to identify, for example, whether predictable patterns of audio features characterise genres or scene types in videos.

Dr. Jones also has an interest in areas of cognitive science relating to human-computer interaction and affective computing. A number of his recent student projects have examined cognitive aspects of information seeking and issues of affect in decision-making and in human and machine learning.