CLARIT compound queries and constraint-controlled feedback in TREC-5 Ad-Hoc experiments
Generate an AI Snapshot to get a quick, structured summary of this paper.
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
The mains objective of the TREC-5 ad-hoc experiment is to evaluate a method by which the user can influence rather than perform the selection of feedback documents for an automatic query enhancement, assuming that the active interaction between the user and the system ends with the initial interactive search over the target data.
Abstract
The mains objective of the TREC-5 ad-hoc experiment is to evaluate a method by which the user can influence rather than perform the selection of feedback documents for an automatic query enhancement, assuming that the active interaction between the user and the system ends with the initial interactive search over the target data. CLARIT ad-hoc experiments represent a continuation and further refinement of the study on constraints controlled feedback but this year they performed ad-hoc experiments with the CLARIT commercial system equipped with a GUI that fully supports the user interactive generation of CLARIT compound queries, a natural language query supplemented by Boolean type constraints. The second objective of the study is to determine how effective user specified constraints are in facilitating the final ranking of documents in order to achieve a higer front-end precision. The official submissions, CLTHES and CLCLUS, explore the use of constraints to control both the automatic feedback and the final ranking of documents. They include in the analysis a new experimental feature of the CLARIT system, namely manual query expansion using pre-computed concept clusters from the target database