Saturday, April 19, 2014
   
Text Size
New ranking methods
Research papers on new search ranking methods and concepts

DocumentsDate added

Order by : Name | Date | Hits [ Ascendant ]
"In view of the recent progress in the field of internet search engines, there is a growing need for mechanisms to evaluate the performance of these useful and popular tools. So far, the vast majority of researchers have relied on the informationretrieval metrics of “precision” and “recall” that quantify the occurrence of “hits” and “misses” in the returned list of documents. What they fail to do is to measure the quality of the ranking that the search engine has provided. This paper wants to rectify the situation. We discuss the issue in some detail, and then propose a new mechanism that we believe is better suited for our needs."
"This tutorial provides an overview on recent advances made in ranking and selection (R&S) for selecting the best simulated system and discusses challenges that still exist in the field. We focus on indifference-zone R&S procedures that provide a guaranteed probability of correct selection when the best system is at least a user-specified amount better than the other systems."
"In this work, we investigate the idea of characterizing the documents and the queries belonging to a given query log with the goal of improving algorithms for detecting spam, both at the document level and at the query leve"l.
"The explosive growth and the widespread accessibility of the Web has led to surge of research activity in the area of information retrieval on the World Wide Web. The seminal papers of Kleinberg [31], and Brin and Page [9] introduced Link Analysis Ranking, where hyperlink structures are used to determine the relative authority of aWeb page, and produce improved algorithms for the ranking of Web search results. In this paper we work within the hubs and authorities framework defined by Kleinberg [31] and we propose new families of algorithms. Two of the algorithms we propose use a Bayesian approach, as opposed to the usual algebraic and graph theoretic approaches. We also introduce a theoretical framework for the study of Link Analysis Ranking algorithms. The framework allows for the definition of specific properties of Link Analysis Ranking algorithms, as well as for comparing different algorithms. We study the properties of the algorithms that we define, and we provide an axiomatic characterization of the INDEGREE heuristic, where each node is ranked according to the number of incoming links. We conclude the paper with an extensive experimental evaluation. We study the quality of the algorithms, and we examine how the existence of different structures in the graphs affect their performance"
"We investigate using gradient descent methods for learning ranking functions; we propose a simple probabilistic cost function, and we introduce RankNet, an implementation of these ideas using a neural network to model the underlying ranking function. We present test results on toy data and on data from a commercial internet search engine."
"Learning to rank is a new statistical learning technology on creating a ranking model for sorting objects. The technology has been successfully applied to web search, and is becoming one of the key machineries for building search engines. Existing approaches to learning to rank, however, did not consider the cases in which there exists relationship between the objects to be ranked, despite of the fact that such situations are very common in practice."
"The past twenty years has seen a rapid growth of interest in stochastic search algorithms, particularly those inspired by natural processes in physics and biology. Impressive results have been demonstrated on complex practical optimisation problems and related search applications taken from a variety of fields, but the theoretical understanding of these algorithms remains weak." 
"The paper deals with the linguistic problem of fully automatic grouping of semantically related words. We discuss the measures of semantic relatedness of basic word forms and describe the treatment of collocations. Next we present the procedure of hierarchical clustering of a very large number of semantically related words and give examples of the resulting partitioning of data in the form of dendrogram. Finally we show a form of the output presentation that facilitates the inspection of the resulting word clusters".
Page 1 of 2
Please update your Flash Player to view content.
Restore Default Settings