Learning to Rank: From Pairwise Approach to Listwise Approach

Proceedings of the 25th …, 2008

This paper aims to conduct a study on the listwise approach to learning to rank. The listwise approach learns a ranking function by taking individual lists as instances and minimizing a loss function defined on the predicted list and the ground-truth list. Existing work on the approach mainly focused on the development of new algorithms; methods such as RankCosine and ListNet have been proposed and good performances by them have been observed. Unfortunately, the underlying theory was not sufficiently studied so far. To amend the problem, this paper proposes conducting theoretical analysis of learning to rank algorithms through investigations on the properties of the loss functions, including consistency, soundness, continuity, differentiability, convexity, and efficiency. A sufficient condition on consistency for ranking is given, which seems to be the first such result obtained in related research. The paper then conducts analysis on three loss functions: likelihood loss, cosine loss, and cross entropy loss. The latter two were used in RankCosine and ListNet. The use of the likelihood loss leads to the development of

Technologies is edited by Graeme Hirst of the University of Toronto. The series consists of 50-to 150-page monographs on topics relating to natural language processing, computational linguistics, information retrieval, and spoken language understanding. Emphasis is on important new techniques, on new applications, and on topics that combine two or more HLT subfields.

Proceedings of the 22nd international conference on Machine learning - ICML '05, 2005

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.

2007

Ranking problem is becoming important in many fields, especially in information retrieval (IR). Many machine learning techniques have been proposed for ranking problem, such as RankSVM, RankBoost, and RankNet. Among them, RankNet, which is based on a probabilistic ranking framework, is leading to promising results and has been applied to a commercial Web search engine. In this paper we conduct further study on the probabilistic ranking framework and provide a novel loss function named fidelity loss for measuring loss of ranking. The fidelity loss not only inherits effective properties of the probabilistic ranking framework in RankNet, but possesses new properties that are helpful for ranking. This includes the fidelity loss obtaining zero for each document pair, and having a finite upper bound that is necessary for conducting query-level normalization. We also propose an algorithm named FRank based on a generalized additive model for the sake of minimizing the fidelity loss and learning an effective ranking function. We evaluated the proposed algorithm for two datasets: TREC dataset and real Web search dataset. The experimental results show that the proposed FRank algorithm outperforms other learning-based ranking methods on both conventional IR problem and Web searching.

Proceeding of the 18th ACM conference on Information and knowledge management - CIKM '09, 2009

Ranking is a key problem in many information retrieval (IR) applications, such as document retrieval and collaborative filtering. In this paper, we address the issue of learning to rank in document retrieval. Learning-based methods, such as RankNet, RankSVM, and RankBoost, try to create ranking functions automatically by using some training data. Recently, several learning to rank methods have been proposed to directly optimize the performance of IR applications in terms of various evaluation measures. They undoubtedly provide statistically significant improvements over conventional methods; however, from the viewpoint of decision-making, most of them do not minimize the Bayes risk of the IR system. In an attempt to fill this research gap, we propose a novel framework that directly optimizes the Bayes risk related to the ranking accuracy in terms of the IR evaluation measures. The results of experiments on the LETOR collections demonstrate that the framework outperforms several existing methods in most cases.

Proceeding of the 17th …, 2008

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. For example, in web search, given a query certain relationships usually exist among the the retrieved documents, e.g., URL hierarchy, similarity, etc., and sometimes it is necessary to utilize the information in ranking of the documents. This paper addresses the issue and formulates it as a novel learning problem, referred to as, 'learning to rank relational objects'. In the new learning task, the ranking model is defined as a function of not only the contents (features) of objects but also the relations between objects. The paper further focuses on one setting of the learning problem in which the way of using relation information is predetermined. It formalizes the learning task as an optimization problem in the setting. The paper then proposes a new method to perform the optimization task, particularly an implementation based on SVM. Experimental results show that the proposed method outperforms the baseline methods for two ranking tasks (Pseudo Relevance Feedback and Topic Distillation) in web search, indicating that the proposed method can indeed make effective use of relation information and content information in ranking.

Information retrieval, 2010

LETOR is a benchmark collection for the research on learning to rank for information retrieval, released by Microsoft Research Asia. In this paper, we describe the details of the LETOR collection and show how it can be used in different kinds of researches. Specifically, we describe how the document corpora and query sets in LETOR are selected, how the documents are sampled, how the learning features and meta information are extracted, and how the datasets are partitioned for comprehensive evaluation. We then compare several state-of-the-art learning to rank algorithms on LETOR, report their ranking performances, and make discussions on the results. After that, we discuss possible new research topics that can be supported by LETOR, in addition to algorithm comparison. We hope that this paper can help people to gain deeper understanding of LETOR, and enable more interesting research projects on learning to rank and related topics.

IEEE Transactions on Neural Networks, 2000

Relevance ranking consists in sorting a set of objects with respect to a given criterion. However, in personalized retrieval systems, the relevance criteria may usually vary among different users and may not be predefined. In this case, ranking algorithms that adapt their behaviour from users' feedbacks must be devised. Two main approaches are proposed in the literature for learning to rank: the use of a scoring function, learned by examples, that evaluates a feature-based representation of each object yielding an absolute relevance score; a pairwise approach, where a preference function is learned to determine the object that has to be ranked first in a given pair. In this paper, we present a preference learning method for learning to rank. A neural network, the Comparative Neural Network (CmpNN), is trained from examples to approximate the comparison function for a pair of objects. The CmpNN adopts a particular architecture designed to implement the symmetries naturally present in a preference function. The learned preference function can be embedded as the comparator into a classical sorting algorithm to provide a global ranking of a set of objects. To improve the ranking performances, an active-learning procedure is devised, that aims at selecting the most informative patterns in the training set. The proposed algorithm is evaluated on the LETOR dataset showing promising performances in comparison with other state of the art algorithms.

Perceptron is a classic online algorithm for learning a classification function. In this paper, we provide a novel extension of the perceptron algorithm to the learning to rank problem in information retrieval. We consider popular listwise performance measures such as Normalized Discounted Cumulative Gain (NDCG) and Average Precision (AP). We propose a novel family of listwise, large margin ranking surrogates, which are adaptable to NDCG and AP measures and derive a perceptron-like algorithm using these surrogates. Exploiting a self-bounding property of the proposed surrogates, we provide a guarantee on the cumulative NDCG (or AP) induced loss incurred by our perceptron-like algorithm. We show that, if there exists a perfect oracle ranker which can correctly rank, with some margin, each instance in an online sequence, the cumulative NDCG (or AP) induced loss of perceptron algorithm on that sequence is bounded by a constant, irrespective of the length of the sequence. This result is a learning to rank analogue of Novikoff's convergence theorem for the classification perceptron. However, our perceptron like algorithm for learning to rank has two drawbacks. First, unlike classification perceptron, the prediction at each round depends on a learning rate parameter. Second, the perceptron loss bound does not match our established lower bound on the cumulative loss achievable by any deterministic online algorithm. We propose a second perceptron like algorithm which achieves the lower bound and is independent of the learning rate parameter. However, our second algorithm does not adapt to different ranking measures, does not possess the listwise property and does not perform well on real world datasets. Experiments on simulated datasets corroborate our theoretical results and demonstrate competitive performance on large industrial benchmark datasets.

2008

The problem of relevance ranking consists of sorting a set of objects with respect to a given criterion. Since users may prefer different relevance criteria, the ranking algorithms should be adaptable to the user needs. Two main approaches exist in literature for the task of learning to rank: 1) a score function, learned by examples, which evaluates the properties of each object yielding an absolute relevance value that can be used to order the objects or 2) a pairwise approach, where a "preference function" is learned using pairs of objects to define which one has to be ranked first. In this paper, we present SortNet, an adaptive ranking algorithm which orders objects using a neural network as a comparator. The neural network training set provides examples of the desired ordering between pairs of items and it is constructed by an iterative procedure which, at each iteration, adds the most informative training examples. Moreover, the comparator adopts a connectionist architecture that is particularly suited for implementing a preference function. We also prove that such an architecture has the universal approximation property and can implement a wide class of functions. Finally, the proposed algorithm is evaluated on the LETOR dataset showing promising performances in comparison with other state of the art algorithms.