Metrics for assessing sets of subtopics

Filip Radlinski, Martin Szummer, Nick Craswell

Abstract: To evaluate the diversity of search results, test collections have been developed that identify multiple intents for each query. Intents are the different meanings or facets that should be covered in a search results list. This means that topic development involves proposing a set of intents for each query. We propose four measurable properties of query-to-intent mappings, allowing for more principled topic development for such test collections.