Open University Expert Search

 

 

 Try it!

 

 

 

 

 

1. What is it?

OU Expert Search (OUES) is a document content based indexing and search engine, and has a Google-like search interface, but OUES finds people instead of documents. Given a search query topic such as "java programming", OUES finds a ranked list of experts from current OU employees who have expertise on the search query topic. Documents supporting expertise are presented and expert list is integrated with the OU staff directory.

2. Who's doing it?

There are many information retrieval groups in the world are working on this now. In KMi, Dr. Jianhan Zhu, Dr. Dawei Song, Prof. Marc Eisenstadt, Prof. Enrico Motta, and Prof. Stefan Ruger are working on OUES. In TREC (Text REtrieval Conference) 2006 expert search task, KMi was ranked top among all 23 internationally renowned research groups across all major Information Retrieval measures.

3. How does it work?

OUES indexes documents on the OU intranet periodically for expert search. Given a search query, OUES estimates the relevance of a person' expertise on the query topic based on the relevance of documents to the topic and the relevance of the person to the topic in these relevant documents.

4. Why is it significant?

Expert finding is very important in the distributed environment like the OU where students and media liaisons are looking for experts over the distance. But it is hard to construct a complete expertise database for OU and such database gets out of date quickly. On the other hand, OUES can provide speedy, most up to date, and accurate expertise information at zero or very low cost.

5. What are the downsides?

People may prefer not to be found, but we can add a privacy filter and ensure that this information is only accessible on the OU Intranet.

6. Where is it going?

We are working on a smoothly integrated search system consisting of OUES and current Ultraseek powered OU intranet search. We are also working on integrating OUES with current OU expertise database and domain knowledge.

7. What are the implications for teaching and learning?

OUES can help teaching and learning by providing advanced expertise search service. For example, (i) students could use OUES to find subject experts (who have granted 'visibility'); (ii) OU course designers can use OUES to assemble a group of experts to work on a course; (iii) on-site experts can be identified to respond to subject-specific enquiries

8. Contacts

Prof. Marc Eisenstadt, Dr. Jianhan Zhu

9. Further reading and resources

  1. TREC: Text Retrieval Conference http://trec.nist.gov/

 

2.       TREC Enterprise Track Wiki site http://www.ins.cwi.nl/projects/trec-ent/wiki/index.php/Main_Page

 

3.       W3C Corpus used for Enterprise Search task http://research.microsoft.com/users/nickcr/w3c-summary.html

 

4.       TREC official publications (2006 proceedings will appear soon) http://trec.nist.gov/pubs.html

 

5.       Jianhan Zhu's publication page: http://kmi.open.ac.uk/people/jianhan/publication.htm

 

10. OUES Screenshot:

 

 

 

 

 

 

 

 

Kmowledge Media Institute

The Open University