1 code implementation • 28 Nov 2023 • Romain Deffayet, Thibaut Thonet, Dongyoon Hwang, Vassilissa Lehoux, Jean-Michel Renders, Maarten de Rijke
Simulators can provide valuable insights for researchers and practitioners who wish to improve recommender systems, because they allow one to easily tweak the experimental setup in which recommender systems operate, and as a result lower the cost of identifying general trends and uncovering novel findings about the candidate methods.
1 code implementation • 4 Nov 2023 • Shiguang Wu, Xin Xin, Pengjie Ren, Zhumin Chen, Jun Ma, Maarten de Rijke, Zhaochun Ren
CSRec contains a teacher module that generates high-quality and confident soft labels and a student module that acts as the target recommender and is trained on the combination of dense, soft labels and sparse, one-hot labels.
1 code implementation • 19 Oct 2023 • Olivier Sprangers, Wander Wadman, Sebastian Schelter, Maarten de Rijke
We implement our sparse hierarchical loss function within an existing forecasting model at bol, a large European e-commerce platform, resulting in an improved forecasting performance of 2% at the product level.
1 code implementation • 18 Oct 2023 • Hengran Zhang, Ruqing Zhang, Jiafeng Guo, Maarten de Rijke, Yixing Fan, Xueqi Cheng
We argue that, rather than relevance, for FV we need to focus on the utility that a claim verifier derives from the retrieved evidence.
1 code implementation • 15 Oct 2023 • Zihan Wang, Ziqi Zhao, Zhumin Chen, Pengjie Ren, Maarten de Rijke, Zhaochun Ren
To address this limitation, recent studies enable generalization to an unseen target domain with only a few labeled examples using data augmentation techniques.
1 code implementation • 18 Sep 2023 • Maria Heuss, Daniel Cohen, Masoud Mansoury, Maarten de Rijke, Carsten Eickhoff
Prior work on bias mitigation often assumes that ranking scores, which correspond to the utility that a document holds for a user, can be accurately determined.
no code implementations • 29 Aug 2023 • Jiangui Chen, Ruqing Zhang, Jiafeng Guo, Maarten de Rijke, Wei Chen, Yixing Fan, Xueqi Cheng
We put forward a novel Continual-LEarner for generatiVE Retrieval (CLEVER) model and make two major contributions to continual learning for GR: (i) To encode new documents into docids with low computational cost, we present Incremental Product Quantization, which updates a partial quantization codebook according to two adaptive thresholds; and (ii) To memorize new documents for querying without forgetting previous knowledge, we propose a memory-augmented learning mechanism, to form meaningful connections between old and new documents.
no code implementations • 19 Aug 2023 • Yu-An Liu, Ruqing Zhang, Jiafeng Guo, Maarten de Rijke, Wei Chen, Yixing Fan, Xueqi Cheng
The AREA task is meant to trick DR models into retrieving a target document that is outside the initial set of candidate documents retrieved by the DR model in response to a query.
no code implementations • 5 Aug 2023 • Ali Vardasbi, Maarten de Rijke, Fernando Diaz, Mostafa Dehghani
With group bias, the utility of the sensitive groups is under-estimated, hence, without correcting for this bias, a supposedly fair ranking is not truly fair.
1 code implementation • 2 Aug 2023 • Ming Li, Mozhdeh Ariannezhad, Andrew Yates, Maarten de Rijke
In next basket recommendation (NBR), it is useful to distinguish between repeat items, i. e., items that a user has consumed before, and explore items, i. e., items that a user has not consumed before.
1 code implementation • 8 Jul 2023 • Weiwei Sun, Hengyi Cai, Hongshen Chen, Pengjie Ren, Zhumin Chen, Maarten de Rijke, Zhaochun Ren
To provide feasible answers to an ambiguous question, one approach is to directly predict all valid answers, but this can struggle with balancing relevance and diversity.
1 code implementation • 15 Jun 2023 • Gabriel Bénédict, Olivier Jeunen, Samuele Papa, Samarth Bhargav, Daan Odijk, Maarten de Rijke
In this paper we propose RecFusion, which comprise a set of diffusion models for recommendation.
no code implementations • 26 May 2023 • Sami Jullien, Romain Deffayet, Jean-Michel Renders, Paul Groth, Maarten de Rijke
Successful applications of distributional reinforcement learning with quantile regression prompt a natural question: can we use other statistics to represent the distribution of returns?
1 code implementation • 22 May 2023 • Vaishali Pal, Andrew Yates, Evangelos Kanoulas, Maarten de Rijke
Recent advances in tabular question answering (QA) with large language models are constrained in their coverage and only answer questions over a single table.
1 code implementation • 18 May 2023 • Chuan Meng, Negar Arabzadeh, Mohammad Aliannejadi, Maarten de Rijke
The QPP task is to predict the retrieval quality of a search system for a query without relevance judgments.
no code implementations • 17 May 2023 • Zihan Wang, Kai Zhao, Yongquan He, Zhumin Chen, Pengjie Ren, Maarten de Rijke, Zhaochun Ren
Recent work on knowledge graph completion (KGC) focused on learning embeddings of entities and relations in knowledge graphs.
1 code implementation • 9 May 2023 • Xin Xin, Xiangyuan Liu, Hanbing Wang, Pengjie Ren, Zhumin Chen, Jiahuan Lei, Xinlei Shi, Hengliang Luo, Joemon Jose, Maarten de Rijke, Zhaochun Ren
Recommender systems that learn from implicit feedback often use large volumes of a single type of implicit user feedback, such as clicks, to enhance the prediction of sparse target behavior such as purchases.
1 code implementation • 1 May 2023 • Fatemeh Sarvi, Ali Vardasbi, Mohammad Aliannejadi, Sebastian Schelter, Maarten de Rijke
We therefore propose an outlier-aware click model that accounts for both outlier and position bias, called outlier-aware position-based model ( OPBM).
1 code implementation • 28 Apr 2023 • Yu-An Liu, Ruqing Zhang, Jiafeng Guo, Maarten de Rijke, Wei Chen, Yixing Fan, Xueqi Cheng
In this paper, we focus on a more general type of perturbation and introduce the topic-oriented adversarial ranking attack task against NRMs, which aims to find an imperceptible perturbation that can promote a target document in ranking for a group of queries with the same topic.
1 code implementation • 28 Apr 2023 • Jiangui Chen, Ruqing Zhang, Jiafeng Guo, Maarten de Rijke, Yiqun Liu, Yixing Fan, Xueqi Cheng
Learning task-specific retrievers that return relevant contexts at an appropriate level of semantic granularity, such as a document retriever, passage retriever, sentence retriever, and entity retriever, may help to achieve better performance on the end-to-end task.
1 code implementation • 26 Apr 2023 • Shashank Gupta, Harrie Oosterhuis, Maarten de Rijke
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.
2 code implementations • 19 Apr 2023 • Romain Deffayet, Philipp Hager, Jean-Michel Renders, Maarten de Rijke
We prove that debiasedness is a necessary condition for recovering unbiased and consistent relevance scores and for the invariance of click prediction under covariate shift.
1 code implementation • 4 Mar 2023 • Yujie Lin, Chenyang Wang, Zhumin Chen, Zhaochun Ren, Xin Xin, Qiang Yan, Maarten de Rijke, Xiuzhen Cheng, Pengjie Ren
STEAM first corrects an input item sequence by adjusting the misclicked and/or missed items.
no code implementations • 20 Jan 2023 • Romain Deffayet, Thibaut Thonet, Jean-Michel Renders, Maarten de Rijke
Our findings suggest that representation learning using generative models is a promising direction towards generalizable RL-based slate recommendation.
1 code implementation • 12 Jan 2023 • Mariya Hendriksen, Svitlana Vakulenko, Ernst Kuiper, Maarten de Rijke
Additionally, we select two scene-centric datasets, and three object-centric datasets, and determine the relative performance of the selected models on these datasets.
1 code implementation • 4 Jan 2023 • Yujie Lin, Zhumin Chen, Zhaochun Ren, Chenyang Wang, Qiang Yan, Maarten de Rijke, Xiuzhen Cheng, Pengjie Ren
To address the limitation of sequential recommenders with side information, we define a way to fuse side information and alleviate the problem of missing side information by proposing a unified task, namely the missing information imputation (MII), which randomly masks some feature fields in a given sequence of items, including item IDs, and then forces a predictive model to recover them.
no code implementations • 3 Jan 2023 • Romain Deffayet, Thibaut Thonet, Jean-Michel Renders, Maarten de Rijke
In this paper, we argue that the paradigm commonly adopted for offline evaluation of sequential recommender systems is unsuitable for evaluating reinforcement learning-based recommenders.
1 code implementation • 20 Dec 2022 • Weiwei Sun, Zhengliang Shi, Shen Gao, Pengjie Ren, Maarten de Rijke, Zhaochun Ren
MixCL effectively reduces the hallucination of LMs in conversations and achieves the highest performance among LM-based dialogue agents in terms of relevancy and factuality.
1 code implementation • 11 Dec 2022 • Yougang Lyu, Piji Li, Yechang Yang, Maarten de Rijke, Pengjie Ren, Yukun Zhao, Dawei Yin, Zhaochun Ren
We also propose a dynamic negative sampling strategy to capture the dynamic influence of biases by employing a bias-only model to dynamically select the most similar biased negative samples.
no code implementations • 17 Sep 2022 • Sanne Vrijenhoek, Gabriel Bénédict, Mateo Gutierrez Granada, Daan Odijk, Maarten de Rijke
In traditional recommender system literature, diversity is often seen as the opposite of similarity, and typically defined as the distance between identified topics, categories or word models.
1 code implementation • 14 Sep 2022 • Chen Wu, Ruqing Zhang, Jiafeng Guo, Wei Chen, Yixing Fan, Maarten de Rijke, Xueqi Cheng
A ranking model is said to be Certified Top-$K$ Robust on a ranked list when it is guaranteed to keep documents that are out of the top $K$ away from the top $K$ under any attack.
1 code implementation • 19 Aug 2022 • Ali Vardasbi, Maarten de Rijke, Mostafa Dehghani
Using this result, we propose to train two parallel instances of a linear model, initialized with different random seeds, and use their intersection as a signal to detect overfitting.
no code implementations • 6 Jul 2022 • Ana Lucic, Sheeraz Ahmad, Amanda Furtado Brinhosa, Vera Liao, Himani Agrawal, Umang Bhatt, Krishnaram Kenthapadi, Alice Xiang, Maarten de Rijke, Nicholas Drabowski
In this paper, we report on ongoing work regarding (i) the development of an AI system for flagging and explaining low-quality medical images in real-time, (ii) an interview study to understand the explanation needs of stakeholders using the AI system at OurCompany, and, (iii) a longitudinal user study design to examine the effect of including explanations on the workflow of the technicians in our clinics.
1 code implementation • 24 Jun 2022 • Zihan Wang, Na Huang, Fei Sun, Pengjie Ren, Zhumin Chen, Hengliang Luo, Maarten de Rijke, Zhaochun Ren
To address the above limitations, we propose a Debiasing Learning for Membership Inference Attacks against recommender systems (DL-MIA) framework that has four main components: (1) a difference vector generator, (2) a disentangled encoder, (3) a weight estimator, and (4) an attack model.
no code implementations • 30 May 2022 • Sami Jullien, Mozhdeh Ariannezhad, Paul Groth, Maarten de Rijke
We frame inventory restocking as a new reinforcement learning task that exhibits stochastic behavior conditioned on the agent's actions, making the environment partially observable.
Distributional Reinforcement Learning
reinforcement-learning
+1
1 code implementation • 25 May 2022 • Maria Heuss, Fatemeh Sarvi, Maarten de Rijke
In this work, we discuss how to approach fairness of exposure in cases where the policy contains rankings of which, due to inter-item dependencies, we cannot reliably estimate the exposure distribution.
1 code implementation • 10 May 2022 • Jin Huang, Harrie Oosterhuis, Bunyamin Cetinkaya, Thijs Rood, Maarten de Rijke
In response to these shortcomings, we reproduce and expand on the existing comparison of attention-based state encoders (1) in the publicly available debiased RL4Rec SOFA simulator with (2) a different RL method, (3) more state encoders, and (4) a different dataset.
1 code implementation • 5 May 2022 • Shaojie Jiang, Ruqing Zhang, Svitlana Vakulenko, Maarten de Rijke
The cross-entropy objective has proved to be an all-purpose training objective for autoregressive language models (LMs).
1 code implementation • 28 Apr 2022 • Maurits Bleeker, Andrew Yates, Maarten de Rijke
We add an additional decoder to the contrastive ICR framework, to reconstruct the input caption in a latent space of a general-purpose sentence encoder, which prevents the image and caption encoder from suppressing predictive features.
1 code implementation • 28 Apr 2022 • Ali Vardasbi, Fatemeh Sarvi, Maarten de Rijke
Different from PL, where pointwise logits are used as the distribution parameters, in PPG pairwise inversion probabilities together with a reference permutation construct the distribution.
no code implementations • 26 Apr 2022 • Clemencia Siro, Mohammad Aliannejadi, Maarten de Rijke
What is the influence of user experience on the user satisfaction rating of TDS as opposed to, or in addition to, utility?
1 code implementation • 15 Apr 2022 • Maartje ter Hoeve, Julia Kiseleva, Maarten de Rijke
Motivated from these two angles, we propose a new task: summarization with graphical elements, and we verify that these summaries are helpful for a critical mass of people.
1 code implementation • dialdoc (ACL) 2022 • Vaishali Pal, Evangelos Kanoulas, Maarten de Rijke
In this work, we study parameter-efficient abstractive QA in encoder-decoder models over structured tabular data and unstructured textual data using only 1. 5% additional parameters for each modality.
no code implementations • 4 Apr 2022 • Chen Wu, Ruqing Zhang, Jiafeng Guo, Maarten de Rijke, Yixing Fan, Xueqi Cheng
We focus on the decision-based black-box attack setting, where the attackers cannot directly get access to the model information, but can only query the target model to obtain the rank positions of the partial retrieved list.
2 code implementations • 2 Apr 2022 • Weiwei Sun, Shuyu Guo, Shuo Zhang, Pengjie Ren, Zhumin Chen, Maarten de Rijke, Zhaochun Ren
Employing existing user simulators to evaluate TDSs is challenging as user simulators are primarily designed to optimize dialogue policies for TDSs and have limited evaluation capabilities.
1 code implementation • 14 Feb 2022 • Maurits Bleeker, Maarten de Rijke
Recent progress in metric learning has given rise to new loss functions that outperform the triplet loss on tasks such as image retrieval and representation learning.
1 code implementation • 21 Dec 2021 • Fatemeh Sarvi, Maria Heuss, Mohammad Aliannejadi, Sebastian Schelter, Maarten de Rijke
We formalize outlierness in a ranking, show that outliers are present in realistic datasets, and present the results of an eye-tracking study, showing that users scanning order and the exposure of items are influenced by the presence of outliers.
1 code implementation • 21 Dec 2021 • Mariya Hendriksen, Maurits Bleeker, Svitlana Vakulenko, Nanne van Noord, Ernst Kuiper, Maarten de Rijke
One aspect of this data is a category tree that is being used in search and recommendation.
1 code implementation • 6 Dec 2021 • Olivier Sprangers, Sebastian Schelter, Maarten de Rijke
However, these methods require a large number of parameters to be learned, which imposes high memory requirements on the computational resources for training such models.
1 code implementation • 24 Nov 2021 • Jin Huang, Harrie Oosterhuis, Maarten de Rijke
We theoretically show that in a dynamic scenario in which both the selection bias and user preferences are dynamic, existing debiasing methods are no longer unbiased.
no code implementations • 1 Nov 2021 • Ana Lucic, Maurits Bleeker, Sami Jullien, Samarth Bhargav, Maarten de Rijke
In this work, we explain the setup for a technical, graduate-level course on Fairness, Accountability, Confidentiality, and Transparency in Artificial Intelligence (FACT-AI) at the University of Amsterdam, which teaches FACT-AI concepts through the lens of reproducibility.
1 code implementation • 29 Sep 2021 • Ming Li, Sami Jullien, Mozhdeh Ariannezhad, Maarten de Rijke
We propose a set of metrics that measure the repeat/explore ratio and performance of NBR models.
1 code implementation • 1 Sep 2021 • Guojun Yan, Jiahuan Pei, Pengjie Ren, Zhaochun Ren, Xin Xin, Huasheng Liang, Maarten de Rijke, Zhumin Chen
(1) there is no dataset with large-scale medical dialogues that covers multiple medical services and contains fine-grained medical labels (i. e., intents, actions, slots, values), and (2) there is no set of established benchmarks for MDSs for multi-domain, multi-service medical dialogues.
1 code implementation • 24 Aug 2021 • Gabriel Bénédict, Vincent Koops, Daan Odijk, Maarten de Rijke
We propose a loss function, sigmoidF1, which is an approximation of the F1 score that (1) is smooth and tractable for stochastic gradient descent, (2) naturally approximates a multilabel metric, and (3) estimates label propensities and label counts.
1 code implementation • 19 Aug 2021 • Ali Vardasbi, Maarten de Rijke, Ilya Markov
Affine correction (AC) is a generalization of IPS that corrects for position bias and trust bias.
1 code implementation • ACL 2021 • Yangjun Zhang, Pengjie Ren, Maarten de Rijke
HMCEval casts dialogue evaluation as a sample assignment problem, where we need to decide to assign a sample to a human or a machine for evaluation.
1 code implementation • In2Writing (ACL) 2022 • Nikos Voskarides, Edgar Meij, Sabrina Sauer, Maarten de Rijke
Given an incomplete narrative that specifies a main event and a context, we aim to retrieve news articles that discuss relevant events that would enable the continuation of the narrative.
1 code implementation • ACL 2021 • Zhongkun Liu, Pengjie Ren, Zhumin Chen, Zhaochun Ren, Maarten de Rijke, Ming Zhou
Conversational Question Simplification (CQS) aims to simplify self-contained questions into conversational ones by incorporating some conversational characteristics, e. g., anaphora and ellipsis.
1 code implementation • 29 Jun 2021 • Shanshan Wang, Pengjie Ren, Zhumin Chen, Zhaochun Ren, Huasheng Liang, Qiang Yan, Evangelos Kanoulas, Maarten de Rijke
We seek to improve the performance for both frequent and rare ICD codes by using a contrastive graph-based EHR coding framework, CoGraph, which re-casts EHR coding as a few-shot learning task.
no code implementations • 23 Jun 2021 • Muyang Ma, Pengjie Ren, Zhumin Chen, Zhaochun Ren, Huasheng Liang, Jun Ma, Maarten de Rijke
By doing so, the SR model is able to learn how to identify common and unique user preferences, and thereby do better user preference extraction and representation.
1 code implementation • 3 Jun 2021 • Olivier Sprangers, Sebastian Schelter, Maarten de Rijke
We propose Probabilistic Gradient Boosting Machines (PGBM), a method to create probabilistic predictions with a single ensemble of decision trees in a computationally efficient manner.
1 code implementation • 18 May 2021 • Pengjie Ren, Zhongkun Liu, Xiaomeng Song, Hongtao Tian, Zhumin Chen, Zhaochun Ren, Maarten de Rijke
(2) We release a benchmark dataset, called wizard of search engine (WISE), which allows for comprehensive and in-depth research on all aspects of CIS.
1 code implementation • 13 May 2021 • Dongdong Li, Zhaochun Ren, Pengjie Ren, Zhumin Chen, Miao Fan, Jun Ma, Maarten de Rijke
We propose an end-to-end variational reasoning approach to medical dialogue generation.
no code implementations • 13 May 2021 • Zihan Wang, Hongye Song, Zhaochun Ren, Pengjie Ren, Zhumin Chen, Xiaozhong Liu, Hongsong Li, Maarten de Rijke
First, contract elements are far more fine-grained than named entities, which hinders the transfer of extractors.
Cross-Domain Named Entity Recognition
named-entity-recognition
+3
1 code implementation • 8 May 2021 • Weiwei Sun, Shuo Zhang, Krisztian Balog, Zhaochun Ren, Pengjie Ren, Zhumin Chen, Maarten de Rijke
The purpose of the task is to increase the evaluation power of user simulations and to make the simulation more human-like.
1 code implementation • 30 Apr 2021 • Ziming Li, Julia Kiseleva, Maarten de Rijke
The proposed backward reasoning step pushes the model to produce more informative and coherent content because the forward generation step's output is used to infer the dialogue context in the backward direction.
no code implementations • 14 Apr 2021 • Svitlana Vakulenko, Evangelos Kanoulas, Maarten de Rijke
These datasets were collected to inform different dialogue-based tasks including conversational search.
1 code implementation • 16 Feb 2021 • Jiahuan Pei, Pengjie Ren, Maarten de Rijke
We find that CoMemNN is able to enrich user profiles effectively, which results in an improvement of 3. 06% in terms of response selection accuracy compared to state-of-the-art methods.
1 code implementation • 11 Feb 2021 • Harrie Oosterhuis, Maarten de Rijke
We introduce the Generalization and Specialization (GENSPEC) algorithm, a robust feature-based counterfactual LTR method that pursues per-query memorization when it is safe to do so.
1 code implementation • 5 Feb 2021 • Ana Lucic, Maartje ter Hoeve, Gabriele Tolomei, Maarten de Rijke, Fabrizio Silvestri
In this work, we propose a method for generating CF explanations for GNNs: the minimal perturbation to the input (graph) data such that the prediction changes.
no code implementations • 23 Jan 2021 • Chongming Gao, Wenqiang Lei, Xiangnan He, Maarten de Rijke, Tat-Seng Chua
In this paper, we provide a systematic review of the techniques used in current CRSs.
1 code implementation • 18 Jan 2021 • Qintong Li, Piji Li, Xinyi Li, Zhaochun Ren, Zhumin Chen, Maarten de Rijke
In this paper, we propose the abstractive opinion tagging task, where systems have to automatically generate a ranked list of opinion tags that are based on, but need not occur in, a given set of user-generated reviews.
no code implementations • 16 Dec 2020 • Mariya Hendriksen, Ernst Kuiper, Pim Nauts, Sebastian Schelter, Maarten de Rijke
In this paper, we focus on purchase prediction for both anonymous and identified sessions on an e-commerce platform.
no code implementations • NAACL 2022 • Maartje ter Hoeve, Julia Kiseleva, Maarten de Rijke
Motivated by our findings, we present ways to mitigate this mismatch in future research on automatic summarization: we propose research directions that impact the design, the development and the evaluation of automatically generated summaries.
1 code implementation • 8 Dec 2020 • Harrie Oosterhuis, Maarten de Rijke
With the introduction of the intervention-aware estimator, we aim to bridge the online/counterfactual LTR division as it is shown to be highly effective in both online and counterfactual scenarios.
no code implementations • 7 Dec 2020 • Svitlana Vakulenko, Vadim Savenkov, Maarten de Rijke
How can we better understand the mechanisms behind multi-turn information seeking dialogues?
1 code implementation • 1 Dec 2020 • Muyang Ma, Pengjie Ren, Zhumin Chen, Zhaochun Ren, Lifan Zhao, Jun Ma, Maarten de Rijke
One of the key challenges in cross-domain sequential recommendation is to grasp and transfer the flow of information from multiple domains so as to promote recommendations in all domains.
1 code implementation • Findings of the Association for Computational Linguistics 2020 • Ziming Li, Sungjin Lee, Baolin Peng, Jinchao Li, Julia Kiseleva, Maarten de Rijke, Shahin Shayandeh, Jianfeng Gao
Reinforcement learning methods have emerged as a popular choice for training an efficient and effective dialogue policy.
no code implementations • CIKM 2020 • Zhiqiang Pan, Fei Cai, Wanyu Chen, Honghui Chen, Maarten de Rijke
The proposed SGNN-HN applies a star graph neural network (SGNN) to model the complex transition relationship between items in an ongoing session.
Ranked #1 on
Session-Based Recommendations
on yoochoose1/64
1 code implementation • Findings of the Association for Computational Linguistics 2020 • Ziming Li, Julia Kiseleva, Maarten de Rijke
Then, the traditional multi-label classification solution for dialogue policy learning is extended by adding dense layers to improve the dialogue agent performance.
1 code implementation • 24 Aug 2020 • Ali Vardasbi, Harrie Oosterhuis, Maarten de Rijke
Our main contribution is a new estimator based on affine corrections: it both reweights clicks and penalizes items displayed on ranks with high trust bias.
no code implementations • 21 Aug 2020 • Yangjun Zhang, Pengjie Ren, Maarten de Rijke
In this paper, we define the task of Malevolent Dialogue Response Detection and Classification (MDRDC).
1 code implementation • 7 Aug 2020 • Phillip Lippe, Pengjie Ren, Hinda Haned, Bart Voorn, Maarten de Rijke
Instead of generating a response from scratch, P2-Net generates system responses by paraphrasing template-based responses.
1 code implementation • 3 Aug 2020 • Anton Steenvoorden, Emanuele Di Gloria, Wanyu Chen, Pengjie Ren, Maarten de Rijke
Users prefer diverse recommendations over homogeneous ones.
1 code implementation • 24 Jul 2020 • Harrie Oosterhuis, Maarten de Rijke
LogOpt turns the counterfactual approach - which is indifferent to the logging policy - into an online approach, where the algorithm decides what rankings to display.
1 code implementation • 20 Jul 2020 • Fatemeh Sarvi, Nikos Voskarides, Lois Mooiman, Sebastian Schelter, Maarten de Rijke
As recent learning to match methods have made important advances in bridging the vocabulary gap for these traditional IR areas, we investigate their potential in the context of product search.
no code implementations • 7 Jul 2020 • Yang Fang, Xiang Zhao, Yifan Chen, Weidong Xiao, Maarten de Rijke
We propose a self-supervised pre-training and fine-tuning framework, PF-HIN, to capture the features of a heterogeneous information network.
no code implementations • 19 Jun 2020 • Ziming Li, Julia Kiseleva, Alekh Agarwal, Maarten de Rijke, Ryen W. White
Effective optimization is essential for real-world interactive systems to provide a satisfactory user experience in response to changing user behavior.
no code implementations • 25 May 2020 • Ali Vardasbi, Maarten de Rijke, Ilya Markov
Unbiased CLTR requires click propensities to compensate for the difference between user clicks and true relevance of search results via IPS.
no code implementations • 25 May 2020 • Svitlana Vakulenko, Evangelos Kanoulas, Maarten de Rijke
The ability to engage in mixed-initiative interaction is one of the core requirements for a conversational search system.
1 code implementation • 24 May 2020 • Nikos Voskarides, Dan Li, Pengjie Ren, Evangelos Kanoulas, Maarten de Rijke
Context from the conversational history can be used to arrive at a better expression of the current turn query, defined as the task of query resolution.
1 code implementation • 21 May 2020 • Rolf Jagerman, Maarten de Rijke
Counterfactual Learning to Rank (LTR) algorithms learn a ranking model from logged user interactions, often collected using a production system.
1 code implementation • 18 May 2020 • Harrie Oosterhuis, Maarten de Rijke
We prove that the policy-aware estimator is unbiased if every relevant item has a non-zero probability to appear in the top-k ranking.
no code implementations • 9 May 2020 • Zhiqiang Pan, Fei Cai, Yanxiang Ling, Maarten de Rijke
We employ a modified self-attention mechanism to estimate item importance in a session, which is then used to predict user's long-term preference.
no code implementations • LREC 2020 • Chuan Wu, Evangelos Kanoulas, Maarten de Rijke, Wei Lu
To support research on entity salience, we present a new dataset, the WikiNews Salience dataset (WN-Salience), which can be used to benchmark tasks such as entity salience detection and salient entity linking.
1 code implementation • 29 Apr 2020 • Pengjie Ren, Zhumin Chen, Zhaochun Ren, Evangelos Kanoulas, Christof Monz, Maarten de Rijke
In this paper, we address the problem of answering complex information needs by conversing conversations with search engines, in the sense that users can express their queries in natural language, and directly receivethe information they need from a short system response in a conversational manner.
1 code implementation • 7 Apr 2020 • Ziming Li, Sungjin Lee, Baolin Peng, Jinchao Li, Julia Kiseleva, Maarten de Rijke, Shahin Shayandeh, Jianfeng Gao
Reinforcement Learning (RL) methods have emerged as a popular choice for training an efficient and effective dialogue policy.
1 code implementation • 26 Mar 2020 • Shaojie Jiang, Thomas Wolf, Christof Monz, Maarten de Rijke
We hypothesize that the deeper reason is that in the training corpora, there are hard tokens that are more difficult for a generative model to learn than others and, once learning has finished, hard tokens are still under-learned, so that repetitive generations are more likely to happen.
1 code implementation • 2 Feb 2020 • Rolf Jagerman, Ilya Markov, Maarten de Rijke
Our experiments using text classification and document retrieval confirm the above by comparing SEA (and a boundless variant called BSEA) to online and offline learning methods for contextual bandit problems.
no code implementations • 27 Jan 2020 • Maartje ter Hoeve, Robert Sim, Elnaz Nouri, Adam Fourney, Maarten de Rijke, Ryen W. White
Our contributions are three-fold: (1) We first present a survey to understand the space of document-centered assistance and the capabilities people expect in this scenario.
1 code implementation • 8 Dec 2019 • Maurits Bleeker, Maarten de Rijke
We introduce the bidirectional Scene Text Transformer (Bi-STET), a novel bidirectional STR method with a single decoder for bidirectional text decoding.
no code implementations • 1 Dec 2019 • Chang Li, Haoyun Feng, Maarten de Rijke
Relevance ranking aims at building a ranked list sorted in decreasing order of item relevance, while result diversification focuses on generating a ranked list of items that covers a broad range of topics.
1 code implementation • 27 Nov 2019 • Ana Lucic, Harrie Oosterhuis, Hinda Haned, Maarten de Rijke
Model interpretability has become an important problem in machine learning (ML) due to the increased effect that algorithmic decisions have on humans.
2 code implementations • 19 Nov 2019 • Jiahuan Pei, Pengjie Ren, Christof Monz, Maarten de Rijke
We propose a novel mixture-of-generators network (MoGNet) for DRG, where we assume that each token of a response is drawn from a mixture of distributions.
no code implementations • 10 Nov 2019 • Joris Baan, Maartje ter Hoeve, Marlies van der Wees, Anne Schuth, Maarten de Rijke
Finally, we find that relative positions heads seem integral to summarization performance and persistently remain after pruning.
no code implementations • 22 Oct 2019 • Yujie Lin, Pengjie Ren, Zhumin Chen, Zhaochun Ren, Dongxiao Yu, Jun Ma, Maarten de Rijke, Xiuzhen Cheng
Given a user, we first obtain a collaborative vector by collecting useful information with a collaborative memory module.
no code implementations • 12 Oct 2019 • Anna Sepliarskaia, Julia Kiseleva, Maarten de Rijke
We conclude that existing metrics of disentanglement were created to reflect different characteristics of disentanglement and do not satisfy two basic desirable properties: (1) assign a high score to representations that are disentangled according to the definition; and (2) assign a low score to representations that are entangled according to the definition.
no code implementations • 6 Oct 2019 • Wenchao Sun, Muyang Ma, Pengjie Ren, Yujie Lin, Zhumin Chen, Zhaochun Ren, Jun Ma, Maarten de Rijke
We study sequential recommendation in a particularly challenging context, in which multiple individual users share asingle account (i. e., they have a shared account) and in which user behavior is available in multiple domains (i. e., recommendations are cross-domain).
no code implementations • 26 Sep 2019 • Jianming Zheng, Fei Cai, Honghui Chen, Maarten de Rijke
We introduce the concept of interaction and propose a two-perspective interaction representation, that encapsulates a local and a global interaction representation.
1 code implementation • 27 Aug 2019 • Wanyu Chen, Pengjie Ren, Fei Cai, Maarten de Rijke
Then, we design an Intent-aware Diversity Promoting (IDP) loss to supervise the learning of the IIM module and force the model to take recommendation diversity into consideration during training.
1 code implementation • 26 Aug 2019 • Pengjie Ren, Zhumin Chen, Christof Monz, Jun Ma, Maarten de Rijke
Given a conversational context and background knowledge, we first learn a topic transition vector to encode the most likely text fragments to be used in the next response, which is then used to guide the local KS at each decoding timestamp.
no code implementations • 24 Aug 2019 • Yujie Lin, Pengjie Ren, Zhumin Chen, Zhaochun Ren, Jun Ma, Maarten de Rijke
FARM improves visual understanding by incorporating the supervision of generation loss, which we hypothesize to be able to better encode aesthetic information.
1 code implementation • 19 Aug 2019 • Svitlana Vakulenko, Javier David Fernandez Garcia, Axel Polleres, Maarten de Rijke, Michael Cochez
We propose a novel approach for complex KGQA that uses unsupervised message passing, which propagates confidence scores obtained by parsing an input question and matching terms in the knowledge graph to a set of possible answers.
1 code implementation • 18 Aug 2019 • Chuan Meng, Pengjie Ren, Zhumin Chen, Christof Monz, Jun Ma, Maarten de Rijke
In this paper, we propose a Reference-aware Network (RefNet) to address the two issues.
1 code implementation • 17 Jul 2019 • Ana Lucic, Hinda Haned, Maarten de Rijke
Given a large error, MC-BRP determines (1) feature values that would result in a reasonable prediction, and (2) general trends between each feature and the target, both based on Monte Carlo simulations.
no code implementations • 16 Jul 2019 • Harrie Oosterhuis, Rolf Jagerman, Maarten de Rijke
Through randomization the effect of different types of bias can be removed from the learning process.
2 code implementations • 15 Jul 2019 • Rolf Jagerman, Harrie Oosterhuis, Maarten de Rijke
At the moment, two methodologies for dealing with bias prevail in the field of LTR: counterfactual methods that learn from historical data and model user behavior to deal with biases; and online methods that perform interventions to deal with bias but use no explicit user models.
no code implementations • 12 Jul 2019 • Alexandra Olteanu, Jean Garcia-Gathright, Maarten de Rijke, Michael D. Ekstrand
The proceedings list for the program of FACTS-IR 2019, the Workshop on Fairness, Accountability, Confidentiality, Transparency, and Safety in Information Retrieval held at SIGIR 2019.
1 code implementation • 10 Jul 2019 • Jiahuan Pei, Pengjie Ren, Maarten de Rijke
We propose a neural Modular Task-oriented Dialogue System(MTDS) framework, in which a few expert bots are combined to generate the response for a given dialogue context.
3 code implementations • 8 Jul 2019 • Wanyu Chen, Fei Cai, Honghui Chen, Maarten de Rijke
Deep feature learning extracts feature representations of users and items with a deep learning architecture based on a user-item rating matrix.
no code implementations • 4 Jul 2019 • Ana Lucic, Hinda Haned, Maarten de Rijke
Understanding how "black-box" models arrive at their predictions has sparked significant interest from both within and outside the AI community.
no code implementations • 1 Jul 2019 • Joris Baan, Maartje ter Hoeve, Marlies van der Wees, Anne Schuth, Maarten de Rijke
We investigate whether distributions calculated by different attention heads in a transformer architecture can be used to improve transparency in the task of abstractive summarization.
no code implementations • 16 Jun 2019 • Jiahuan Pei, Arent Stienstra, Julia Kiseleva, Maarten de Rijke
Obtaining key information from a complex, long dialogue context is challenging, especially when different sources of information are available, e. g., the user's utterances, the system's responses, and results retrieved from a knowledge base (KB).
1 code implementation • 16 Jun 2019 • Yangjun Zhang, Pengjie Ren, Maarten de Rijke
The latter generate responses thatare natural but not necessarily effective in leveraging background knowledge.
no code implementations • 29 May 2019 • Chang Li, Maarten de Rijke
We consider the problem of identifying the K most attractive items and propose cascading non-stationary bandits, an online learning variant of the cascading model, where a user browses a ranked list from top to bottom and clicks on the first attractive item.
no code implementations • 7 Mar 2019 • Bram van den Akker, Ilya Markov, Maarten de Rijke
The visual appearance of a webpage carries valuable information about its quality and can be used to improve the performance of learning to rank (LTR).
2 code implementations • 25 Feb 2019 • Shaojie Jiang, Pengjie Ren, Christof Monz, Maarten de Rijke
Specifically, we first analyze the influence of the commonly used Cross-Entropy (CE) loss function, and find that the CE loss function prefers high-frequency tokens, which results in low-diversity responses.
no code implementations • 27 Dec 2018 • Svitlana Vakulenko, Kate Revoredo, Claudio Di Ciccio, Maarten de Rijke
Understanding the structure of interaction processes helps us to improve information-seeking dialogue systems.
1 code implementation • 11 Dec 2018 • Chang Li, Ilya Markov, Maarten de Rijke, Masrour Zoghi
Our main finding is that for large-scale Condorcet ranker evaluation problems, MergeDTS outperforms the state-of-the-art dueling bandit algorithms.
1 code implementation • 9 Dec 2018 • Ziming Li, Julia Kiseleva, Maarten de Rijke
The performance of adversarial dialogue generation models relies on the quality of the reward signal produced by the discriminator.
1 code implementation • 6 Dec 2018 • Pengjie Ren, Zhumin Chen, Jing Li, Zhaochun Ren, Jun Ma, Maarten de Rijke
RepeatNet integrates a regular neural recommendation approach in the decoder with a new repeat recommendation mechanism that can choose items from a user's history and recommends them at the right time.
1 code implementation • 12 Oct 2018 • Hosein Azarbonyad, Mostafa Dehghani, Tom Kenter, Maarten Marx, Jaap Kamps, Maarten de Rijke
For measuring topical diversity of text documents, our HiTR approach improves over the state-of-the-art measured on PubMed dataset.
1 code implementation • 22 Sep 2018 • Harrie Oosterhuis, Maarten de Rijke
Instead, its gradient is based on inferring preferences between document pairs from user clicks and can optimize any differentiable model.
no code implementations • WS 2018 • Shaojie Jiang, Maarten de Rijke
Sequence-to-sequence (Seq2Seq) models have been shown to be very effective for response generation.
no code implementations • 23 Jun 2018 • Yujie Lin, Pengjie Ren, Zhumin Chen, Zhaochun Ren, Jun Ma, Maarten de Rijke
The generated comments can be regarded as explanations for the recommendation results.
1 code implementation • 18 Jun 2018 • Chang Li, Maarten de Rijke
Ordinal Regression (OR) aims to model the ordering information between different data categories, which is a crucial topic in multi-label learning.
2 code implementations • 17 Jun 2018 • Svitlana Vakulenko, Maarten de Rijke, Michael Cochez, Vadim Savenkov, Axel Polleres
Conversational systems have become increasingly popular as a way for humans to interact with computers.
no code implementations • 15 Jun 2018 • Chang Li, Branislav Kveton, Tor Lattimore, Ilya Markov, Maarten de Rijke, Csaba Szepesvari, Masrour Zoghi
In this paper, we study the problem of safe online learning to re-rank, where user feedback is used to improve the quality of displayed lists.
no code implementations • 6 Jun 2018 • Sapna Negi, Maarten de Rijke, Paul Buitelaar
We first present an annotation study, and based on our observations propose a formal task definition and annotation procedure for creating benchmark datasets for suggestion mining.
no code implementations • 14 May 2018 • Maartje ter Hoeve, Anne Schuth, Daan Odijk, Maarten de Rijke
There is an increasing demand for algorithms to explain their outcomes.
1 code implementation • 7 May 2018 • Harrie Oosterhuis, Maarten de Rijke
Existing learning to rank methods cannot handle such complex ranking settings as they assume that the display order is known beforehand.
no code implementations • 7 May 2018 • Nikos Voskarides, Edgar Meij, Ridho Reinanda, Abhinav Khaitan, Miles Osborne, Giorgio Stefanoni, Prabhanjan Kambadur, Maarten de Rijke
KG fact contextualization is the task of augmenting a given KG fact with additional and useful KG facts.
1 code implementation • ICML 2018 • Boris Sharchilev, Yury Ustinovsky, Pavel Serdyukov, Maarten de Rijke
We address the problem of finding influential training samples for a particular case of tree ensemble-based models, e. g., Random Forest (RF) or Gradient Boosted Decision Trees (GBDT).
no code implementations • 17 Feb 2018 • Ziming Li, Julia Kiseleva, Alekh Agarwal, Maarten de Rijke
Effective optimization is essential for interactive systems to provide a satisfactory user experience.
no code implementations • ICLR 2018 • Thorsten Joachims, Adith Swaminathan, Maarten de Rijke
We propose a new output layer for deep neural networks that permits the use of logged contextual bandit feedback for training.
1 code implementation • 19 Dec 2017 • Tom Kenter, Maarten de Rijke
We argue that the process of building a representation of the conversation can be framed as a machine reading task, where an automated system is presented with a number of statements about which it should answer questions.
no code implementations • 15 Sep 2017 • Svitlana Vakulenko, Ilya Markov, Maarten de Rijke
In this paper we investigate the affordances of interactive storytelling as a tool to enable exploratory search within the framework of a conversational interface.
4 code implementations • 9 Aug 2017 • Christophe Van Gysel, Maarten de Rijke, Evangelos Kanoulas
We propose the Neural Vector Space Model (NVSM), a method that learns representations of documents in an unsupervised manner for news article retrieval.
1 code implementation • 25 Jul 2017 • Christophe Van Gysel, Maarten de Rijke, Evangelos Kanoulas
We discover how clusterings of experts correspond to committees in organizations, the ability of expert representations to encode the co-author graph, and the degree to which they encode academic rank.
1 code implementation • 24 Jul 2017 • Rolf Jagerman, Julia Kiseleva, Maarten de Rijke
List-wise learning to rank methods are considered to be the state-of-the-art.
no code implementations • 24 Jul 2017 • Mostafa Dehghani, Hosein Azarbonyad, Jaap Kamps, Maarten de Rijke
Deep neural networks have become a primary tool for solving problems in many fields.
no code implementations • 24 Jul 2017 • Dat Tien Nguyen, Shafiq Joty, Basma El Amel Boussaha, Maarten de Rijke
Discussion forums are an important source of information.
no code implementations • 13 Jul 2017 • Tom Kenter, Alexey Borisov, Christophe Van Gysel, Mostafa Dehghani, Maarten de Rijke, Bhaskar Mitra
Machine learning plays a role in many aspects of modern IR systems, and deep learning is applied in all of them.
1 code implementation • 12 Jun 2017 • Christophe Van Gysel, Maarten de Rijke, Evangelos Kanoulas
Unsupervised learning of low-dimensional, semantic representations of words and entities has recently gained attention.
1 code implementation • 15 Jan 2017 • David Graus, Daan Odijk, Maarten de Rijke
We do so by tracking entities that emerge in public discourse, that is, in online text streams such as social media and news streams, before they are incorporated into Wikipedia, which, we argue, can be viewed as an online place for collective memory.
1 code implementation • 3 Jan 2017 • Christophe Van Gysel, Evangelos Kanoulas, Maarten de Rijke
We introduce pyndri, a Python interface to the Indri search engine.
no code implementations • 1 Dec 2016 • Damien Lefortier, Adith Swaminathan, Xiaotao Gu, Thorsten Joachims, Maarten de Rijke
The ability to perform effective off-policy learning would revolutionize the process of building better interactive systems, such as search engines and recommendation systems for e-commerce, computational advertising and news.
no code implementations • 8 Nov 2016 • Kezban Dilek Onal, Ismail Sengor Altingovde, Pinar Karagoz, Maarten de Rijke
We detail the required background and terminology, a taxonomy grouping the rapidly growing body of work in the area, and then survey work on neural models for semantic matching in the context of three tasks: query suggestion, ad retrieval, and document retrieval.
no code implementations • 18 Sep 2016 • Bingbing Jiang, Chang Li, Maarten de Rijke, Xin Yao, Huanhuan Chen
The proposed method, called probabilistic feature selection and classification vector machine (PFCVMLP ), is able to simultaneously select relevant features and samples for classification tasks.
1 code implementation • 2 Sep 2016 • Aleksandr Chuklin, Maarten de Rijke
In this paper we propose a model of user behavior on a SERP that jointly captures click behavior, user attention and satisfaction, the CAS model, and demonstrate that it gives more accurate predictions of user actions and self-reported satisfaction than existing models based on clicks alone.
2 code implementations • 25 Aug 2016 • Christophe Van Gysel, Maarten de Rijke, Evangelos Kanoulas
We introduce a novel latent vector space model that jointly learns the latent representations of words, e-commerce products and a mapping between the two without the need for explicit annotations.
1 code implementation • 23 Aug 2016 • Christophe Van Gysel, Evangelos Kanoulas, Maarten de Rijke
Lexical query modeling has been the leading paradigm for session search.
1 code implementation • 23 Aug 2016 • Christophe Van Gysel, Maarten de Rijke, Marcel Worring
We compare our model to state-of-the-art unsupervised statistical vector space and probabilistic generative approaches.
2 code implementations • ACL 2016 • Tom Kenter, Alexey Borisov, Maarten de Rijke
We present the Siamese Continuous Bag of Words (Siamese CBOW) model, a neural network for efficient estimation of high-quality sentence embeddings.
1 code implementation • 24 May 2016 • Rolf Jagerman, Carsten Eickhoff, Maarten de Rijke
Topic models such as Latent Dirichlet Allocation (LDA) have been widely used in information retrieval for tasks ranging from smoothing and feedback methods to tools for exploratory search and discovery.