KMi Seminars
Expert Finding - Academia vs. Practice
This event took place on Wednesday 09 January 2008 at 11:30

 
Thijs Westerveld Teezir search solutions

With the start of an enterprise search track at TREC in 2005, the search for topical expertise has recently received quite some attention in the academic world. The practical value of an expert finding system is evident. Connecting people to interact and share their knowledge is widely recognised as an important factor in the successful operation of an enterprise or organisation. In this talk I will start to outline the field of expert finding from these two perspectives. I will discuss the
set-up of expert finding task as organized by TREC's expert finding track as well as the practical usefulness of such systems. The second part of the talk will focus on previous work I did in the context of the TREC benchmarks. Typically, expert finding systems follow one of two approaches. Either they build profiles for candidate experts based on the documents associated to them and ranking the profiles, or they create a document ranking aggregate document scores into person scores based on the person-document associations. I will discuss an approach that is somehow in between these two approaches and produces two document rankings, one topic based and one person based. The correlation between topical document ranking and personal document ranking is taken as an indication of the person's expertise. Finally, I will highlight the differences between the academic evaluation of expert finding and the real life situations as we find them in practice.

 
KMi Seminars Event | SSSW 2013, The 10th Summer School on Ontology Engineering and the Semantic Web Journal | 25 years of knowledge acquisition
 

Multimedia and Information Systems is...


Multimedia and Information Systems
Our research is centred around the theme of Multimedia Information Retrieval, ie, Video Search Engines, Image Databases, Spoken Document Retrieval, Music Retrieval, Query Languages and Query Mediation.

We focus on content-based information retrieval over a wide range of data spanning form unstructured text and unlabelled images over spoken documents and music to videos. This encompasses the modelling of human perception of relevance and similarity, the learning from user actions and the up-to-date presentation of information. Currently we are building a research version of an integrated multimedia information retrieval system MIR to be used as a research prototype. We aim for a system that understands the user's information need and successfully links it to the appropriate information sources, be it a report or a TV news clip. This work is guided by the vision that an automated knowledge extraction system ultimately empowers people making efficient use of information sources without the burden of filing data into specialised databases.

Visit the MMIS website