Ranking is at the core of many artificial intelligence (AI) applications, including search engines, recommender systems, etc. Modern ranking systems are often constructed with learning-to-rank (LTR) models built from user behavior signals. While previous studies have demonstrated the effectiveness of using user behavior signals (e.g., clicks) as both features and labels of LTR algorithms, we argue that existing LTR algorithms that indiscriminately treat behavior and non-behavior signals in input features could lead to suboptimal performance in practice. Particularly because user behavior signals often have strong correlations with the ranking objective and can only be collected on items that have already been shown to users, directly using behavior signals in LTR could create an exploitation bias that hurts the system performance in the long run. To address the exploitation bias, we propose EBRank, an empirical Bayes-based uncertainty-aware ranking algorithm. Specifically, to overcome exploitation bias brought by behavior features in ranking models, EBRank uses a sole non-behavior feature based prior model to get a prior estimation of relevance. In the dynamic training and serving of ranking systems, EBRank uses the observed user behaviors to update posterior relevance estimation instead of concatenating behaviors as features in ranking models. Besides, EBRank additionally applies an uncertainty-aware exploration strategy to explore actively, collect user behaviors for empirical Bayesian modeling and improve ranking performance. Experiments on three public datasets show that EBRank is effective, practical and significantly outperforms state-of-the-art ranking algorithms.
Online learning to rank (OLTR) is a sequential decision-making problem where a learning agent selects an ordered list of items and receives feedback through user clicks. Although potential attacks against OLTR algorithms may cause serious losses in real-world applications, little is known about adversarial attacks on OLTR. This paper studies attack strategies against multiple variants of OLTR. Our first result provides an attack strategy against the UCB algorithm on classical stochastic bandits with binary feedback, which solves the key issues caused by bounded and discrete feedback that previous works can not handle. Building on this result, we design attack algorithms against UCB-based OLTR algorithms in position-based and cascade models. Finally, we propose a general attack strategy against any algorithm under the general click model. Each attack algorithm manipulates the learning agent into choosing the target attack item $T-o(T)$ times, incurring a cumulative cost of $o(T)$. Experiments on synthetic and real data further validate the effectiveness of our proposed attack algorithms.
Unsupervised sentence representation learning is one of the fundamental problems in natural language processing with various downstream applications. Recently, contrastive learning has been widely adopted which derives high-quality sentence representations by pulling similar semantics closer and pushing dissimilar ones away. However, these methods fail to capture the fine-grained ranking information among the sentences, where each sentence is only treated as either positive or negative. In many real-world scenarios, one needs to distinguish and rank the sentences based on their similarities to a query sentence, e.g., very relevant, moderate relevant, less relevant, irrelevant, etc. In this paper, we propose a novel approach, RankCSE, for unsupervised sentence representation learning, which incorporates ranking consistency and ranking distillation with contrastive learning into a unified framework. In particular, we learn semantically discriminative sentence representations by simultaneously ensuring ranking consistency between two representations with different dropout masks, and distilling listwise ranking knowledge from the teacher. An extensive set of experiments are conducted on both semantic textual similarity (STS) and transfer (TR) tasks. Experimental results demonstrate the superior performance of our approach over several state-of-the-art baselines.
Deep neural classifiers trained with cross-entropy loss (CE loss) often suffer from poor calibration, necessitating the task of out-of-distribution (OOD) detection. Traditional supervised OOD detection methods require expensive manual annotation of in-distribution and OOD samples. To address the annotation bottleneck, we introduce SELFOOD, a self-supervised OOD detection method that requires only in-distribution samples as supervision. We cast OOD detection as an inter-document intra-label (IDIL) ranking problem and train the classifier with our pairwise ranking loss, referred to as IDIL loss. Specifically, given a set of in-distribution documents and their labels, for each label, we train the classifier to rank the softmax scores of documents belonging to that label to be higher than the scores of documents that belong to other labels. Unlike CE loss, our IDIL loss function reaches zero when the desired confidence ranking is achieved and gradients are backpropagated to decrease probabilities associated with incorrect labels rather than continuously increasing the probability of the correct label. Extensive experiments with several classifiers on multiple classification datasets demonstrate the effectiveness of our method in both coarse- and fine-grained settings.
The identification of important nodes with strong propagation capabilities in road networks is a vital topic in urban planning. Existing methods for evaluating the importance of nodes in traffic networks only consider topological information and traffic volumes, the diversity of the traffic characteristics in road networks, such as the number of lanes and average speed of road segments, is ignored, thus limiting their performance. To solve this problem, we propose a graph learning-based framework (MGL2Rank) that integrates the rich characteristics of road networks to rank the importance of nodes. This framework comprises an embedding module containing a sampling algorithm (MGWalk) and an encoder network to learn the latent representations for each road segment. MGWalk utilizes multigraph fusion to capture the topology of road networks and establish associations between road segments based on their attributes. The obtained node representation is then used to learn the importance ranking of the road segments. Finally, a synthetic dataset is constructed for ranking tasks based on the regional road network of Shenyang City, and the ranking results on this dataset demonstrate the effectiveness of our method. The data and source code for MGL2Rank are available at https://github.com/iCityLab/MGL2Rank.
Estimating position bias is a well-known challenge in Learning to Rank (L2R). Click data in e-commerce applications, such as targeted advertisements and search engines, provides implicit but abundant feedback to improve personalized rankings. However, click data inherently includes various biases like position bias. Based on the position-based click model, Result Randomization and Regression Expectation-Maximization algorithm (REM) have been proposed to estimate position bias, but they require various paired observations of (item, position). In real-world scenarios of advertising, marketers frequently display advertisements in a fixed pre-determined order, which creates difficulties in estimation due to the limited availability of various pairs in the training data, resulting in a sparse dataset. We propose a variant of the REM that utilizes item embeddings to alleviate the sparsity of (item, position). Using a public dataset and internal carousel advertisement click dataset, we empirically show that item embedding with Latent Semantic Indexing (LSI) and Variational Auto-Encoder (VAE) improves the accuracy of position bias estimation and the estimated position bias enhances Learning to Rank performance. We also show that LSI is more effective as an embedding creation method for position bias estimation.
Query-focused meeting summarization(QFMS) aims to generate a specific summary for the given query according to the meeting transcripts. Due to the conflict between long meetings and limited input size, previous works mainly adopt extract-then-summarize methods, which use extractors to simulate binary labels or ROUGE scores to extract utterances related to the query and then generate a summary. However, the previous approach fails to fully use the comparison between utterances. To the extractor, comparison orders are more important than specific scores. In this paper, we propose a Ranker-Generator framework. It learns to rank the utterances by comparing them in pairs and learning from the global orders, then uses top utterances as the generator's input. We show that learning to rank utterances helps to select utterances related to the query effectively, and the summarizer can benefit from it. Experimental results on QMSum show that the proposed model outperforms all existing multi-stage models with fewer parameters.
The goal of unbiased learning to rank (ULTR) is to leverage implicit user feedback for optimizing learning-to-rank systems. Among existing solutions, automatic ULTR algorithms that jointly learn user bias models (i.e., propensity models) with unbiased rankers have received a lot of attention due to their superior performance and low deployment cost in practice. Despite their theoretical soundness, the effectiveness is usually justified under a weak logging policy, where the ranking model can barely rank documents according to their relevance to the query. However, when the logging policy is strong, e.g., an industry-deployed ranking policy, the reported effectiveness cannot be reproduced. In this paper, we first investigate ULTR from a causal perspective and uncover a negative result: existing ULTR algorithms fail to address the issue of propensity overestimation caused by the query-document relevance confounder. Then, we propose a new learning objective based on backdoor adjustment and highlight its differences from conventional propensity models, which reveal the prevalence of propensity overestimation. On top of that, we introduce a novel propensity model called Logging-Policy-aware Propensity (LPP) model and its distinctive two-step optimization strategy, which allows for the joint learning of LPP and ranking models within the automatic ULTR framework, and actualize the unconfounded propensity estimation for ULTR. Extensive experiments on two benchmarks demonstrate the effectiveness and generalizability of the proposed method.
This monograph takes a step towards promoting the study of efficiency in the era of neural information retrieval by offering a comprehensive survey of the literature on efficiency and effectiveness in ranking, and to a limited extent, retrieval. This monograph was inspired by the parallels that exist between the challenges in neural network-based ranking solutions and their predecessors, decision forest-based learning to rank models, as well as the connections between the solutions the literature to date has to offer. We believe that by understanding the fundamentals underpinning these algorithmic and data structure solutions for containing the contentious relationship between efficiency and effectiveness, one can better identify future directions and more efficiently determine the merits of ideas. We also present what we believe to be important research directions in the forefront of efficiency and effectiveness in retrieval and ranking.
This paper describes the approach of the THUIR team at the COLIEE 2023 Legal Case Entailment task. This task requires the participant to identify a specific paragraph from a given supporting case that entails the decision for the query case. We try traditional lexical matching methods and pre-trained language models with different sizes. Furthermore, learning-to-rank methods are employed to further improve performance. However, learning-to-rank is not very robust on this task. which suggests that answer passages cannot simply be determined with information retrieval techniques. Experimental results show that more parameters and legal knowledge contribute to the legal case entailment task. Finally, we get the third place in COLIEE 2023. The implementation of our method can be found at https://github.com/CSHaitao/THUIR-COLIEE2023.
Legal case retrieval techniques play an essential role in modern intelligent legal systems. As an annually well-known international competition, COLIEE is aiming to achieve the state-of-the-art retrieval model for legal texts. This paper summarizes the approach of the championship team THUIR in COLIEE 2023. To be specific, we design structure-aware pre-trained language models to enhance the understanding of legal cases. Furthermore, we propose heuristic pre-processing and post-processing approaches to reduce the influence of irrelevant messages. In the end, learning-to-rank methods are employed to merge features with different dimensions. Experimental results demonstrate the superiority of our proposal. Official results show that our run has the best performance among all submissions. The implementation of our method can be found at https://github.com/CSHaitao/THUIR-COLIEE2023.
Recent work has shown that standard training via empirical risk minimization (ERM) can produce models that achieve high accuracy on average but low accuracy on underrepresented groups due to the prevalence of spurious features. A predominant approach to tackle this group robustness problem minimizes the worst group error (akin to a minimax strategy) on the training data, hoping it will generalize well on the testing data. However, this is often suboptimal, especially when the out-of-distribution (OOD) test data contains previously unseen groups. Inspired by ideas from the information retrieval and learning-to-rank literature, this paper first proposes to use Discounted Cumulative Gain (DCG) as a metric of model quality for facilitating better hyperparameter tuning and model selection. Being a ranking-based metric, DCG weights multiple poorly-performing groups (instead of considering just the group with the worst performance). As a natural next step, we build on our results to propose a ranking-based training method called Discounted Rank Upweighting (DRU), which differentially reweights a ranked list of poorly-performing groups in the training data to learn models that exhibit strong OOD performance on the test data. Results on several synthetic and real-world datasets highlight the superior generalization ability of our group-ranking-based (akin to soft-minimax) approach in selecting and learning models that are robust to group distributional shifts.
Since its inception, the field of unbiased learning to rank (ULTR) has remained very active and has seen several impactful advancements in recent years. This tutorial provides both an introduction to the core concepts of the field and an overview of recent advancements in its foundations along with several applications of its methods. The tutorial is divided into four parts: Firstly, we give an overview of the different forms of bias that can be addressed with ULTR methods. Secondly, we present a comprehensive discussion of the latest estimation techniques in the ULTR field. Thirdly, we survey published results of ULTR in real-world applications. Fourthly, we discuss the connection between ULTR and fairness in ranking. We end by briefly reflecting on the future of ULTR research and its applications. This tutorial is intended to benefit both researchers and industry practitioners who are interested in developing new ULTR solutions or utilizing them in real-world applications.
Counterfactual learning to rank (CLTR) relies on exposure-based inverse propensity scoring (IPS), a LTR-specific adaptation of IPS to correct for position bias. While IPS can provide unbiased and consistent estimates, it often suffers from high variance. Especially when little click data is available, this variance can cause CLTR to learn sub-optimal ranking behavior. Consequently, existing CLTR methods bring significant risks with them, as naively deploying their models can result in very negative user experiences. We introduce a novel risk-aware CLTR method with theoretical guarantees for safe deployment. We apply a novel exposure-based concept of risk regularization to IPS estimation for LTR. Our risk regularization penalizes the mismatch between the ranking behavior of a learned model and a given safe model. Thereby, it ensures that learned ranking models stay close to a trusted model, when there is high uncertainty in IPS estimation, which greatly reduces the risks during deployment. Our experimental results demonstrate the efficacy of our proposed method, which is effective at avoiding initial periods of bad performance when little data is available, while also maintaining high performance at convergence. For the CLTR field, our novel exposure-based risk minimization method enables practitioners to adopt CLTR methods in a safer manner that mitigates many of the risks attached to previous methods.
Bandit algorithms for online learning to rank (OLTR) problems often aim to maximize long-term revenue by utilizing user feedback. From a practical point of view, however, such algorithms have a high risk of hurting user experience due to their aggressive exploration. Thus, there has been a rising demand for safe exploration in recent years. One approach to safe exploration is to gradually enhance the quality of an original ranking that is already guaranteed acceptable quality. In this paper, we propose a safe OLTR algorithm that efficiently exchanges one of the items in the current ranking with an item outside the ranking (i.e., an unranked item) to perform exploration. We select an unranked item optimistically to explore based on Kullback-Leibler upper confidence bounds (KL-UCB) and safely re-rank the items including the selected one. Through experiments, we demonstrate that the proposed algorithm improves long-term regret from baselines without any safety violation.
User interaction data is an important source of supervision in counterfactual learning to rank (CLTR). Such data suffers from presentation bias. Much work in unbiased learning to rank (ULTR) focuses on position bias, i.e., items at higher ranks are more likely to be examined and clicked. Inter-item dependencies also influence examination probabilities, with outlier items in a ranking as an important example. Outliers are defined as items that observably deviate from the rest and therefore stand out in the ranking. In this paper, we identify and introduce the bias brought about by outlier items: users tend to click more on outlier items and their close neighbors. To this end, we first conduct a controlled experiment to study the effect of outliers on user clicks. Next, to examine whether the findings from our controlled experiment generalize to naturalistic situations, we explore real-world click logs from an e-commerce platform. We show that, in both scenarios, users tend to click significantly more on outlier items than on non-outlier items in the same rankings. We show that this tendency holds for all positions, i.e., for any specific position, an item receives more interactions when presented as an outlier as opposed to a non-outlier item. We conclude from our analysis that the effect of outliers on clicks is a type of bias that should be addressed in ULTR. We therefore propose an outlier-aware click model that accounts for both outlier and position bias, called outlier-aware position-based model ( OPBM). We estimate click propensities based on OPBM ; through extensive experiments performed on both real-world e-commerce data and semi-synthetic data, we verify the effectiveness of our outlier-aware click model. Our results show the superiority of OPBM against baselines in terms of ranking performance and true relevance estimation.
Many re-ranking strategies in search systems rely on stochastic ranking policies, encoded as Doubly-Stochastic (DS) matrices, that satisfy desired ranking constraints in expectation, e.g., Fairness of Exposure (FOE). These strategies are generally two-stage pipelines: \emph{i)} an offline re-ranking policy construction step and \emph{ii)} an online sampling of rankings step. Building a re-ranking policy requires repeatedly solving a constrained optimization problem, one for each issued query. Thus, it is necessary to recompute the optimization procedure for any new/unseen query. Regarding sampling, the Birkhoff-von-Neumann decomposition (BvND) is the favored approach to draw rankings from any DS-based policy. However, the BvND is too costly to compute online. Hence, the BvND as a sampling solution is memory-consuming as it can grow as $\gO(N\, n^2)$ for $N$ queries and $n$ documents. This paper offers a novel, fast, lightweight way to predict fair stochastic re-ranking policies: Constrained Meta-Optimal Transport (CoMOT). This method fits a neural network shared across queries like a learning-to-rank system. We also introduce Gumbel-Matching Sampling (GumMS), an online sampling approach from DS-based policies. Our proposed pipeline, CoMOT + GumMS, only needs to store the parameters of a single model, and it generalizes to unseen queries. We empirically evaluated our pipeline on the TREC 2019 and 2020 datasets under FOE constraints. Our experiments show that CoMOT rapidly predicts fair re-ranking policies on held-out data, with a speed-up proportional to the average number of documents per query. It also displays fairness and ranking performance similar to the original optimization-based policy. Furthermore, we empirically validate the effectiveness of GumMS to approximate DS-based policies in expectation.
This paper introduces the approaches we have used to participate in the WSDM Cup 2023 Task 1: Unbiased Learning to Rank. In brief, we have attempted a combination of both traditional IR models and transformer-based cross-encoder architectures. To further enhance the ranking performance, we also considered a series of features for learning to rank. As a result, we won 2nd place on the final leaderboard.
In the stock market, a successful investment requires a good balance between profits and risks. Based on the learning to rank paradigm, stock recommendation has been widely studied in quantitative finance to recommend stocks with higher return ratios for investors. Despite the efforts to make profits, many existing recommendation approaches still have some limitations in risk control, which may lead to intolerable paper losses in practical stock investing. To effectively reduce risks, we draw inspiration from adversarial learning and propose a novel Split Variational Adversarial Training (SVAT) method for risk-aware stock recommendation. Essentially, SVAT encourages the stock model to be sensitive to adversarial perturbations of risky stock examples and enhances the model's risk awareness by learning from perturbations. To generate representative adversarial examples as risk indicators, we devise a variational perturbation generator to model diverse risk factors. Particularly, the variational architecture enables our method to provide a rough risk quantification for investors, showing an additional advantage of interpretability. Experiments on several real-world stock market datasets demonstrate the superiority of our SVAT method. By lowering the volatility of the stock recommendation model, SVAT effectively reduces investment risks and outperforms state-of-the-art baselines by more than 30% in terms of risk-adjusted profits. All the experimental data and source code are available at https://drive.google.com/drive/folders/14AdM7WENEvIp5x5bV3zV_i4Aev21C9g6?usp=sharing.
A well-known problem when learning from user clicks are inherent biases prevalent in the data, such as position or trust bias. Click models are a common method for extracting information from user clicks, such as document relevance in web search, or to estimate click biases for downstream applications such as counterfactual learning-to-rank, ad placement, or fair ranking. Recent work shows that the current evaluation practices in the community fail to guarantee that a well-performing click model generalizes well to downstream tasks in which the ranking distribution differs from the training distribution, i.e., under covariate shift. In this work, we propose an evaluation metric based on conditional independence testing to detect a lack of robustness to covariate shift in click models. We introduce the concept of debiasedness in click modeling and derive a metric for measuring it. In extensive semi-synthetic experiments, we show that our proposed metric helps to predict the downstream performance of click models under covariate shift and is useful in an off-policy model selection setting.