Journal ArticleDOI
When is the probability ranking principle suboptimal
Michael D. Gordon,Peter Lenk +1 more
Reads0
Chats0
TLDR
The probability ranking principle retrieves documents in decreasing order of their predictive probabilities of relevance and can be suboptimal with respect to expected utility when one of these conditions fails to hold.Abstract:
The probability ranking principle retrieves documents in decreasing order of their predictive probabilities of relevance. Gordon and Lenk (1991) demonstrated that this principal is optimal within a signal detection-decision theory framework, and it maximizes the inquirer's expected utility for relevant documents. These results hold under three conditions: calibration, independent assessment of relevance by the inquirer, and certainty about the computed probabilities of relevance. We demonstrate that the probability ranking principle can be suboptimal with respect to expected utility when one of these conditions fails to hold.read more
Citations
More filters
Book
Search Result Diversification
TL;DR: A formal definition of the search result diversification problem is provided and the most successful approaches in the literature for producing and evaluating diversity in multiple search domains are described.
Book ChapterDOI
Using the quantum probability ranking principle to rank interdependent documents
Guido Zuccon,Leif Azzopardi +1 more
TL;DR: In this article, the Quantum Probability Ranking Principle (QPRP) has been proposed, which implicitly captures dependencies between documents through "quantum interference" and leads to improved performance for subtopic retrieval, where novelty and diversity is required.
Using the quantum probability ranking principle to rank interdependent documents
Guido Zuccon,Leif Azzopardi +1 more
TL;DR: It is shown that the application of quantum theory to problems within information retrieval can lead to significant improvements, and the QPRP outperforms other ranking strategies for subtopic retrieval.
Exploration and contextualization through interaction and concepts
TL;DR: This thesis investigates two dimensions along which tools to support humanities researchers in dealing with the flood of information may be improved: by providing richer means of interaction with information systems and developing algorithms that allow discovery of information through relations between concepts.
Posted Content
Due Diligence Failure as a Signal Detection Problem
TL;DR: In this paper, the authors propose that the relative importance acquirers place on these two risks affects how they utilize information obtained during due-diligence, and conduct an experimental study of decision-making in due- diligence.