Rank aggregation methods for the Web

Dwork, Cynthia; Kumar, Ravi; Naor, Moni; Sivakumar, D.

doi:10.1145/371920.372165

Cited by 1,397 publications

(1,347 citation statements)

References 15 publications

Supporting

Mentioning

1,328

Contrasting

Unclassified

Order By: Relevance

“…Furthermore, there are several studies which have discussed blend ranking or rank aggregation [15,16]. But, they targeted the problem of merging the different rankings on the homogeneous set of items, i.e.…”

Section: Related Workmentioning

confidence: 99%

Learning to blend vitality rankings from heterogeneous social networks

Bian

Chang

et al. 2012

Neurocomputing

View full text Add to dashboard Cite

Heterogeneous social network services, such as Facebook and Twitter, have emerged as popular, and often effective channels for Web users to capture updates from their friends. The explosion in popularity of these social network services, however, has created the problem of "information overload". The problem is becoming more severe as more and more users have engaged in more than one social networks simultaneously, each of which usually yields different friend connections and various sources of updates. Thus, it has made necessity to perform effective information filtering to retrieve information really attractive to web users from each of social networks and further blend them into a unified ranking list. In this paper, we introduce the problem of blending vitality rankings from heterogeneous social networks, where vitality denotes all kinds of updates user receives in various social networks. We propose a variety of content, users, and users correlation features for this task. Since vitalities from different social networks are likely to have different sets of features, we employ a divide-and-conquer strategy in order to fully exploit all available features for vitalities from each social network, respectively. Our experimental results, obtained from a large scale evaluation over two popular social networks, demonstrate the effectiveness of our method for putting vitalities that really interest users into higher orders in the blended ranking list. We complement our results with a thorough investigation of the feature importance and model selection with respect to both blending strategy and ranking for each social network.

show abstract

Section: Related Workmentioning

confidence: 99%

Learning to blend vitality rankings from heterogeneous social networks

Bian

Chang

et al. 2012

Neurocomputing

View full text Add to dashboard Cite

show abstract

“…The learning task is to combine these rankings into a complete ranking of all objects in X . A practical application of this setting occurs, e.g., in information retrieval, when different rankings of search results originating from different search engines should be combined into an overall ranking of all retrieved pages [16]. Amongst other things, the learning problem may involve the determination of suitable weights for the information sources (search engines), reflecting their performance or agreement with the preferences of the user [35].…”

Section: Local Aggregation Of Preferencesmentioning

confidence: 99%

Preference Learning: An Introduction

Fürnkranz¹,

Hüllermeier

2010

Preference Learning

106

View full text Add to dashboard Cite

Abstract. This introduction gives a brief overview of the field of preference learning and, along the way, tries to establish a unified terminology. Special emphasis will be put on learning to rank, which is by now one of the most extensively studied problem tasks in preference learning and also prominently represented in this book. We propose a categorization of ranking problems into object ranking, instance ranking, and label ranking. Moreover, we introduce these scenarios in a formal way, discuss different ways in which the learning of ranking functions can be approached, and explain how the contributions collected in this book relate to this categorization. Finally, we also highlight some important applications of preference learning methods.

show abstract

“…The combination of several sources of ranking has been the object of active research in the field of IR [3]. We have adopted the so-called combSUM model, by which the two rankings are merged by a linear combination of the relevance scores:…”

Section: Personalization Effectmentioning

confidence: 99%

“…The specificity of a search is defined as a function spec (q) = f (spec 1 (q), spec 2 (T q ), spec 3 …”

Section: Assessing the Vagueness Of The Searchmentioning

confidence: 99%

“…But a better definition of "personalization impact" can be given in terms of the effective change of position of documents in the ranking. Dwork et al [3] propose the Spearman footrule distance to measure the difference between two search result lists as the average displacement of each document across the rankings. We argue that a more significant measure of impact is whether or not a result will be seen at all by the user.…”

Section: Adjusting Personalization By Impactmentioning

confidence: 99%

See 1 more Smart Citation