550 likes | 694 Views
Relevance Ranking in the Scholarly Domain. Dr. Tamar Sadeh LIBER Conference Tartu, Estonia, June 2012. The top three keys for success. 1. 2. 3. Content. 1. Speed. 2. Relevance Ranking. 3.
E N D
Relevance Ranking in the Scholarly Domain • Dr. Tamar Sadeh • LIBER Conference • Tartu, Estonia, June 2012
Content 1
Speed 2
Relevance is the measure of correspondence between a document and a query as determined by a user Based on Saracevic, 1975
System or algorithmic relevance • Topical or subject relevance • Cognitive relevance or pertinence • Situational relevance or utility • Effective relevance • System or algorithmic relevance • Topical or subject relevance • Cognitive relevance or pertinence • Situational relevance or utility • Affective relevance
The Goal Enhance the Primo relevance ranking algorithm
Setting up a team • Building test environment, tools, and procedures • Defining metrics to evaluate our current success and the improvements we make • Defining measurements to assess the success of the changes, once implemented
Working with researchers • Researchers’ evaluation quantified • Enhancements introduced and checked in the lab, using defined metrics • Enhancements launched and usage patterns monitored • Improvements are introduced on an ongoing basis
abstractauthordatefull text journallanguagetypepublishersubjecttitle citationsdownloadsjournal impactfactoreigenfactorpagerank
academic degree discipline(s)languagelocationprevious selections search history
? broad-topic search currency exact-item search material type narrow-topic search
Narrow-topic query Known-item query Broad-topic query Author-related query
The match: traditional information retrieval methods, adapted to the scholarly environment
1 100 20
? no. of citations; no. of selections; recency; type; peer review
1 20 100
Author-related query, known-item query, broad-topic query… ?