Opportunities for Ph.D. candidates
I have a number of possible vacancies for Ph.D. candidates interested in the following areas:
- Information Retrieval
- Query performance prediction
Can we predict the quality of the result set returned in response to a query? See for example pdf
- Index organization
The inverted index is the key data structure used in information retrieval. How can we optimize its structure with respect to storage and computation?
- Trust
Retrieval results contain documents from a wide variety of sources. Which sources can be trusted? What are the primary sources for particular news items? Are these sources independent and verifiable?
- Query-independent document ranking
Pagerank is an example of a query-independent document ranking. However, there are many collections, e.g. internal company networks, for which hypertext pointers are sparse or unavailable. What other information can be used to infer the importance of documents?
- Signal processing for text retrieval
Text can be considered as equivalent to a one-dimensional temporal signal. What signal processing techniques can be usefully applied to the problem of text retrieval?
- Multimedia content recognition (fingerprinting)
This research examines the problem of identifying music or video from a short example. The key issues are representation and search. This project is expected to develop new algorithms for nearest neighbor search in high dimensional spaces. See for example pdf.
- Web modeling
This research builds is an extension of work on relevance feedback applied to the Web. It is intended to model properties of the Web that can then be exploited by relevance feedback algorithms.
- Stereo Reconstruction
This research looks at the problem of reconstructing synthetic views of a scene given two of more real views. See for example pdf.
Please contact me at ingemar at ieee dot org