Expertise Recommendation: A triangulated approach

Abstract — Recommender systems are becoming increasingly popular as a means for bringing products to the attention of online users. Similarly, they offer a means by which scarce resources in the form of human experts can be identified and accessed. However, if the information in the system is missing, incorrect or obsolete, recommendations will not be followed or even sought in the first place. Relying on individuals to validate and update this information is problematic. To provide automated acquisition and maintenance of information regarding who has expertise and in what areas, we employ data mining techniques. However, data mining will not provide the full picture and thus our outputs are reviewed by the experts themselves, providing a second means of validation. The third part to our triangulated approach is the use of profiles and the gathering of feedback from both searchers and experts to ensure that recommendations provided are satisfactory to both parties