Abstract
The most commonly used performance measures represent the characteristics of a search up to a certain point in the retrieval process. These measures include precision, recall, and Van Rijsbergen’s and Shaw’s E measure. Swets’ E measure, on the other hand, addresses the performance of a particular ranking based on the characteristics of the distributions for relevant and non-relevant documents. It is therefore independent of the stopping point chosen in a particular search. The relationship between ordered sequences, such as lists of ranked documents, may be studied using a number of techniques (Levenstein, 1966; Myers & Miller, 1988), ranging from sophisticated cost-based methods to counting the number of swaps necessary in sorting one sequence so that it becomes the other sequence. How best to study the relationships between sequences is an open question.
There is [a]... great cause of the little advancement of the sciences, which is this: it is impossible to advance properly in the course when the goal is not properly fixed.
—Francis Bacon, 1620
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1998 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Losee, R.M. (1998). The Quality of a Ranking Method. In: Text Retrieval and Filtering. The Information Retrieval Series, vol 3. Springer, Boston, MA. https://doi.org/10.1007/978-1-4615-5705-0_5
Download citation
DOI: https://doi.org/10.1007/978-1-4615-5705-0_5
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4613-7612-5
Online ISBN: 978-1-4615-5705-0
eBook Packages: Springer Book Archive