Maarten de Rijke

Information retrieval

Author: mdr (page 1 of 15)

IJCAI 2019 papers online

Cascading non-stationary bandits: Online learning to rank in the non-stationary cascade model by Chang Li and Maarten de Rijke is online now at this location.

In the paper, we argue that non-stationarity appears in many online applications such as web search and advertising. We study the online learning to rank problem in a non-stationary environment where user preferences change abruptly at an unknown moment in time. 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. We propose two algorithms for solving this non-stationary problem: CascadeDUCB andCascadeSWUCB. We analyze their performance and derive gap-dependent upper bounds on the $n$-step regret of these algorithms. We also establish a lower bound on the regret for cascading non-stationary bandits and show that both algorithms match the lower bound up to a logarithmic factor. Finally, we evaluate their performance on a real-world web search click dataset.

  • Chang Li and Maarten de Rijke. Cascading non-stationary bandits: Online learning to rank in the non-stationary cascade model. In IJCAI 2019: Twenty-Eighth International Joint Conference on Artificial Intelligence, page 2859–2865, August 2019. Bibtex, PDF
    @inproceedings{li-2019-cascading,
    Author = {Li, Chang and de Rijke, Maarten},
    Booktitle = {IJCAI 2019: Twenty-Eighth International Joint Conference on Artificial Intelligence},
    Date-Added = {2019-05-30 22:36:52 +0200},
    Date-Modified = {2019-08-04 15:53:45 +0200},
    Month = {August},
    Pages = {2859--2865},
    Title = {Cascading non-stationary bandits: Online learning to rank in the non-stationary cascade model},
    Year = {2019}}

Other papers and presentations at IJCAI are part of the SCAI workshop:

  • Jiahuan Pei, Arent Stienstra, Julia Kiseleva, and Maarten de Rijke. SEntNet: Source-aware Recurrent Entity Networks for Dialogue Response Selection. In 4th International Workshop on Search-Oriented Conversational AI (SCAI), August 2019. Bibtex, PDF
    @inproceedings{pei-2019-sentnet,
    Author = {Pei, Jiahuan and Stienstra, Arent and Kiseleva, Julia and de Rijke, Maarten},
    Booktitle = {4th International Workshop on Search-Oriented Conversational AI (SCAI)},
    Date-Added = {2019-06-06 11:55:06 +0200},
    Date-Modified = {2019-06-06 11:56:16 +0200},
    Month = {August},
    Title = {SEntNet: Source-aware Recurrent Entity Networks for Dialogue Response Selection},
    Year = {2019}}
  • Yangjun Zhang, Pengjie Ren, and Maarten de Rijke. Improving Background Based Conversation with Context-aware Knowledge Pre-selection. In 4th International Workshop on Search-Oriented Conversational AI (SCAI), August 2019. Bibtex, PDF
    @inproceedings{zhang-2019-improving,
    Author = {Zhang, Yangjun and Ren, Pengjie and de Rijke, Maarten},
    Booktitle = {4th International Workshop on Search-Oriented Conversational AI (SCAI)},
    Date-Added = {2019-06-06 11:53:36 +0200},
    Date-Modified = {2019-06-06 11:55:01 +0200},
    Month = {August},
    Title = {Improving Background Based Conversation with Context-aware Knowledge Pre-selection},
    Year = {2019}}

  • Maarten de Rijke and Pengjie Ren. SERP-based Conversations. In 4th International Workshop on Search-Oriented Conversational AI (SCAI), August 2019. 

SIGIR 2019 papers online

Here’s our harvest for SIGIR 2019, which is about to get started in less than 24 hours:

  • Joris Baan, Maartje ter Hoeve, Marlies van der Wees, Anne Schuth, and Maarten de Rijke. Do Transformer Attention Heads Provide Transparency in Abstractive Summarization?. In FACTS-IR: SIGIR 2019 Workshop on Fairness, Accountability, Confidentiality, Transparency and Safety in Information Retrieval, July 2019. Bibtex, PDF
    @inproceedings{baan-2019-do,
    Author = {Baan, Joris and ter Hoeve, Maartje and van der Wees, Marlies and Schuth, Anne and de Rijke, Maarten},
    Booktitle = {FACTS-IR: SIGIR 2019 Workshop on Fairness, Accountability, Confidentiality, Transparency and Safety in Information Retrieval},
    Date-Added = {2019-05-31 22:24:14 +0200},
    Date-Modified = {2019-05-31 22:38:02 +0200},
    Month = {July},
    Title = {Do Transformer Attention Heads Provide Transparency in Abstractive Summarization?},
    Year = {2019}}
  • Yifan Chen, Pengjie Ren, Yang Wang, and Maarten de Rijke. Bayesian Personalized Feature Interaction Selection for Factorization Machines. In SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval, page 665–674. ACM, July 2019. Bibtex, PDF
    @inproceedings{chen-2019-bayesian,
    Author = {Chen, Yifan and Ren, Pengjie and Wang, Yang and de Rijke, Maarten},
    Booktitle = {SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval},
    Date-Added = {2019-04-14 16:48:31 +0200},
    Date-Modified = {2019-08-02 15:58:43 +0200},
    Month = {July},
    Pages = {665--674},
    Publisher = {ACM},
    Title = {Bayesian Personalized Feature Interaction Selection for Factorization Machines},
    Year = {2019}}
  • Yang Fang, Xiang Zhao, Peixin Huang, Weidong Xiao, and Maarten de Rijke. M-HIN: Complex Embeddings for Heterogeneous Information Networks via Metagraphs. In SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval, page 913–916. ACM, July 2019. Bibtex, PDF
    @inproceedings{fang-2019-m-hin,
    Author = {Fang, Yang and Zhao, Xiang and Huang, Peixin and Xiao, Weidong and de Rijke, Maarten},
    Booktitle = {SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval},
    Date-Added = {2019-04-14 21:00:49 +0200},
    Date-Modified = {2019-08-02 15:59:43 +0200},
    Month = {July},
    Pages = {913--916},
    Publisher = {ACM},
    Title = {M-HIN: Complex Embeddings for Heterogeneous Information Networks via Metagraphs},
    Year = {2019}}
  • Rolf Jagerman, Harrie Oosterhuis, and Maarten de Rijke. To Model or to Intervene: A Comparison of Counterfactual and Online Learning to Rank from User Interactions. In SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval, page 15–24. ACM, July 2019. Bibtex, PDF
    @inproceedings{jagerman-2019-model,
    Author = {Jagerman, Rolf and Oosterhuis, Harrie and de Rijke, Maarten},
    Booktitle = {SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval},
    Date-Added = {2019-04-14 16:45:43 +0200},
    Date-Modified = {2019-08-02 15:58:05 +0200},
    Month = {July},
    Pages = {15--24},
    Publisher = {ACM},
    Title = {To Model or to Intervene: A Comparison of Counterfactual and Online Learning to Rank from User Interactions},
    Year = {2019}}
  • Claudio Lucchese, Franco Maria Nardini, Rama Kumar Pasumarthi, Sebastian Bruch, Michael Bendersky, Xuanhui Wang, Harrie Oosterhuis, Rolf Jagerman, and Maarten de Rijke. Learning to Rank in Theory and Practice: From Gradient Boosting to Neural Networks and Unbiased Learning. In SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval, page 1419–1420, ACM, July 2019. Bibtex, PDF
    @inproceedings{lucchese-2019-learning,
    Address = {ACM},
    Author = {Lucchese, Claudio and Nardini, Franco Maria and Pasumarthi, Rama Kumar and Bruch, Sebastian and Bendersky, Michael and Wang, Xuanhui and Oosterhuis, Harrie and Jagerman, Rolf and de Rijke, Maarten},
    Booktitle = {SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval},
    Date-Added = {2019-05-31 22:47:26 +0200},
    Date-Modified = {2019-08-02 16:00:12 +0200},
    Month = {July},
    Pages = {1419--1420},
    Title = {Learning to Rank in Theory and Practice: From Gradient Boosting to Neural Networks and Unbiased Learning},
    Year = {2019}}
  • Ana Lucic, Hinda Haned, and Maarten de Rijke. Explaining Predictions from Tree-based Boosting Ensembles. In FACTS-IR: SIGIR 2019 Workshop on Fairness, Accountability, Confidentiality, Transparency and Safety in Information Retrieval, July 2019. Bibtex, PDF
    @inproceedings{lucic-2019-explaining,
    Author = {Lucic, Ana and Haned, Hinda and de Rijke, Maarten},
    Booktitle = {FACTS-IR: SIGIR 2019 Workshop on Fairness, Accountability, Confidentiality, Transparency and Safety in Information Retrieval},
    Date-Added = {2019-05-31 22:21:39 +0200},
    Date-Modified = {2019-05-31 22:38:12 +0200},
    Month = {July},
    Title = {Explaining Predictions from Tree-based Boosting Ensembles},
    Year = {2019}}
  • Muyang Ma, Pengjie Ren, Yujie Lin, Zhumin Chen, Jun Ma, and Maarten de Rijke. $\pi$-Net: A Parallel Information-sharing Network for Cross-domain Shared-account Sequential Recommendations. In SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval, page 685–694. ACM, July 2019. Bibtex, PDF
    @inproceedings{ma-2019-pi-net,
    Author = {Ma, Muyang and Ren, Pengjie and Lin, Yujie and Chen, Zhumin and Ma, Jun and de Rijke, Maarten},
    Booktitle = {SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval},
    Date-Added = {2019-04-14 16:49:57 +0200},
    Date-Modified = {2019-08-02 15:59:03 +0200},
    Month = {July},
    Pages = {685--694},
    Publisher = {ACM},
    Title = {$\pi$-Net: A Parallel Information-sharing Network for Cross-domain Shared-account Sequential Recommendations},
    Year = {2019}}
  • Alexandra Olteanu, Jean Garcia-Gathright, Maarten de Rijke, and Michael D. Ekstrand. Workshop on Fairness, Accountability, Confidentiality, Transparency, and Safety in Information Retrieval (FACTS-IR). In SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval, page 1423–1425. ACM, July 2019. Bibtex, PDF
    @inproceedings{olteanu-2019-workshop,
    Author = {Olteanu, Alexandra and Garcia-Gathright, Jean and de Rijke, Maarten and Ekstrand, Michael D.},
    Booktitle = {SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval},
    Date-Added = {2019-05-31 22:51:14 +0200},
    Date-Modified = {2019-08-02 16:00:36 +0200},
    Month = {July},
    Pages = {1423--1425},
    Publisher = {ACM},
    Title = {Workshop on Fairness, Accountability, Confidentiality, Transparency, and Safety in Information Retrieval (FACTS-IR)},
    Year = {2019}}
  • Jiahuan Pei, Pengjie Ren, and Maarten de Rijke. A Modular Task-oriented Dialogue System Using a Neural Mixture-of-Experts. In WCIS: SIGIR 2019 Workshop on Conversational Interaction Systems. ACM, July 2019. Bibtex, PDF
    @inproceedings{pei-2019-modular,
    Author = {Pei, Jiahuan and Ren, Pengjie and de Rijke, Maarten},
    Booktitle = {WCIS: SIGIR 2019 Workshop on Conversational Interaction Systems},
    Date-Added = {2019-06-01 07:18:46 +0200},
    Date-Modified = {2019-07-14 12:24:23 +0200},
    Month = {July},
    Publisher = {ACM},
    Title = {A Modular Task-oriented Dialogue System Using a Neural Mixture-of-Experts},
    Year = {2019}}
  • Taihua Shao, Fei Cai, Honghui Chen, and Maarten de Rijke. Length-adaptive Neural Network for Answer Selection. In SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval, page 869–872. ACM, July 2019. Bibtex, PDF
    @inproceedings{shao-2019-length-adaptive,
    Author = {Shao, Taihua and Cai, Fei and Chen, Honghui and de Rijke, Maarten},
    Booktitle = {SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval},
    Date-Added = {2019-04-14 20:59:02 +0200},
    Date-Modified = {2019-08-02 15:59:20 +0200},
    Month = {July},
    Pages = {869--872},
    Publisher = {ACM},
    Title = {Length-adaptive Neural Network for Answer Selection},
    Year = {2019}}
  • Meirui Wang, Pengjie Ren, Lei Mei, Zhumin Chen, Jun Ma, and Maarten de Rijke. A Collaborative Session-based Recommendation Approach with Parallel Memory Modules. In SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval, page 345–354. ACM, July 2019. Bibtex, PDF
    @inproceedings{wang-2019-collaborative,
    Author = {Wang, Meirui and Ren, Pengjie and Mei, Lei and Chen, Zhumin and Ma, Jun and de Rijke, Maarten},
    Booktitle = {SIGIR 2019: 42nd international ACM SIGIR conference on Research and Development in Information Retrieval},
    Date-Added = {2019-04-14 16:52:16 +0200},
    Date-Modified = {2019-08-02 15:58:23 +0200},
    Month = {July},
    Pages = {345--354},
    Publisher = {ACM},
    Title = {A Collaborative Session-based Recommendation Approach with Parallel Memory Modules},
    Year = {2019}}

UAI 2019 paper on safe online learning to re-rank via implicit click feedback online

BubbleRank: Safe Online Learning to Re-Rank via Implicit Click Feedback by Chang Li, Branislav Kveton, Tor Lattimore, Ilya Markov, Maarten de Rijke, Csaba Szepesvari, and Masrour Zoghi is online now at this location.

In the paper we study the problem of safe on-line learning to re-rank, where user feedback is used to improve the quality of displayed lists. Learning to rank has traditionally been studied in two settings. In the offline setting, rankers are typically learned from relevance labels created by judges. This approach has generally be- come standard in industrial applications of ranking, such as search. However, this approach lacks exploration and thus is limited by the information content of the offline training data. In the online setting, an algorithm can experiment with lists and learn from feedback on them in a sequential fashion. Bandit algorithms are well-suited for this setting but they tend to learn user preferences from scratch, which results in a high initial cost of exploration. This poses an additional challenge of safe exploration in ranked lists. We propose BubbleRank, a bandit algorithm for safe re-ranking that combines the strengths of both the offline and online settings. The algorithm starts with an initial base list and improves it online by gradually exchanging higher-ranked less attractive items for lower-ranked more attractive items. We prove an up- per bound on the n-step regret of BubbleRankthat degrades gracefully with the quality of the initial base list. Our theoretical findings are supported by extensive experiments on a large-scale real-world click dataset.

The paper will be presented at UAI 2019: Conference on Uncertainty in Artificial Intelligence, July 2019.

Using process mining for understanding the structure of interaction processes

Svitlana Vakulenko explains how we have recently used process mining techniques to understand the structure of interaction processes, which will in turn help us to improve information-seeking dialogue systems. We extract a new model of information-seeking dialogues, QRFA, for Query, Request, Feedback, Answer. The QRFA model better reflects conversation flows observed in real information-seeking conversations than models proposed previously. QRFA allows us to identify malfunctioning in dialogue system transcripts as deviations from the expected conversation flow described by the model via conformance analysis.

Read the full post.

Investeer in kennisbasis AI of word een toeschouwer

In een opiniestuk voor NRC Handelsblad en NRC Next beargumenteer ik dat artificiële intelligentie ons leven hoe dan ook zal veranderen en dat Nederland voor de keuze staat om vol mee te doen in de ontwikkeling van AI en het spel mee te bepalen, of om de bank te blijven zitten. Wie niet actief meedoet, heeft geen invloed – niet op het spel en al helemaal niet op de spelregels. Investeer in de AI-kennisbasis, investeer in talent. Kom van de bank!

Het hele stuk is hier te vinden.

Learning to answer questions by taking broader contexts into account

Mostafa Dehghani has posted an explanation of our recent work on TraCRNet (“tracker net”) to learn how to answer questions from multiple, possible long documents. TraCRNet uses the universal transformer and is able to go beyond understanding a set of input documents separately and combine their information in multiple steps. TraCRNet is highly parallellizable and far more robust against noisy input than previous proposals for addressing the question answering task.

See this page for the full post.

FACTS-IR Workshop @ SIGIR 2019

SIGIR 2019 will host a workshop to explore challenges in responsible information retrieval system development and deployment. The focus will be on determining actionable research agendas on five key dimensions of responsible information retrieval: fairness, accountability, confidentiality, transparency, and safety. Rather than just a mini-conference, this workshop will be an event during which participants will also be expected to work. The workshop aims to bring together a diverse set of researchers and practitioners interested in helping to develop the technical research agenda for responsible information retrieval.

The web site for the workshop is live now.

How to optimize ranking systems by directly interacting with users

Harrie Oosterhuis has written an accessible summary of our recent work on pairwise differentiable gradient descent (PDGD), an online learning to rank method that he published at CIKM 2018, with a follow-up paper to come at ECIR 2019 in April. With the introduction of the PDGD algorithm, ranking systems can now be optimized from user interactions far more effectively than previously possible. Additionally, PDGD can also optimize neural models to a greater effect, something previous methods couldn’t do. We expect that the development of ranking systems will benefit from this contribution on the long term. Not only because of improved performance, but also because the possibility to optimize more complex models opens the door to many different possibilities.

See this page for the full post.

WWW 2019 paper on evaluation metrics for web image search online

Grid-based Evaluation Metrics for Web Image Search by Xiaohui Xie, Jiaxin Mao, Yiqun Liu, Maarten de Rijke, Yunqiu Shao, Zixin Ye, Min Zhang, and Shaoping Ma is online now at this location.

Compared to general web search engines, web image search engines display results in a different way. In web image search, results are typically placed in a grid-based manner rather than a sequential result list. In this scenario, users can view results not only in a vertical direction but also in a horizontal direction. Moreover, pagination is usually not (explicitly) supported on image search search engine result pages (SERPs), and users can view results by scrolling down without having to click a “next page” button. These differences lead to different interaction mechanisms and user behavior patterns, which, in turn, create challenges to evaluation metrics that have originally been developed for general web search. While considerable effort has been invested in developing evaluation metrics for general web search, there have been relatively less effort to construct grid-based evaluation metrics.

To inform the development of grid-based evaluation metrics for web image search, we conduct a comprehensive analysis of user behavior so as to uncover how users allocate their attention in a grid-based web image search result interface. We obtain three findings: (1) “Middle bias”: Confirming previous studies, we find that image results in the horizontal middle positions may receive more attention from users than those in the leftmost or rightmost positions. (2) “Slower decay”: Unlike web search, users’ attention does not decrease monotonically or dramatically with the rank position in image search, especially within a row. (3) “Row skipping”: Users may ignore particular rows and directly jump to results at some distance. Motivated by these observations, we propose correspond- ing user behavior assumptions to capture users’ search interaction processes and evaluate their search performance. We show how to derive new metrics from these assumptions and demonstrate that they can be adopted to revise traditional list-based metrics like Discounted Cumulative Gain (DCG) and Rank-Biased Precision (RBP). To show the effectiveness of the proposed grid-based metrics, we compare them against a number of list-based metrics in terms of their correlation with user satisfaction. Our experimental results show that the proposed grid-based evaluation metrics better reflect user satisfaction in web image search.

The paper will be presented at The Web Conference 2019.

WWW 2019 paper on outfit recommendation online

Improving Outfit Recommendation with Co-supervision of Fashion Generation by Yujie Lin, Pengjie Ren, Zhumin Chen, Zhaochun Ren, Jun Ma, and Maarten de Rijke is now available at this location.

The task of fashion recommendation includes two main challenges:visual understanding and visual matching. Visual understanding aims to extract effective visual features. Visual matching aims to model a human notion of compatibility to compute a match between fashion items. Most previous studies rely on recommendation loss alone to guide visual understanding and matching. Although the features captured by these methods describe basic characteristics (e.g., color, texture, shape) of the input items, they are not directly related to the visual signals of the output items (to be recommended). This is problematic because the aesthetic characteristics (e.g., style, design), based on which we can directly infer the output items, are lacking. Features are learned under the recommendation loss alone, where the supervision signal is simply whether the given two items are matched or not.

To address this problem, we propose a neural co-supervision learning framework, called the FAshion Recommendation Machine (FARM). FARM improves visual understanding by incorporating the supervision of generation loss, which we hypothesize to be able to better encode aesthetic information. FARM enhances visual matching by introducing a novel layer-to-layer matching mechanism to fuse aesthetic information more effectively, and meanwhile avoiding paying too much attention to the generation quality and ignoring the recommendation performance.

Extensive experiments on two publicly available datasets show that FARM outperforms state-of-the-art models on outfit recom- mendation, in terms of AUC and MRR. Detailed analyses of gener- ated and recommended items demonstrate that FARM can encode better features and generate high quality images as references to improve recommendation performance.

The paper will be presented at the The Web Conference 2019.

« Older posts

© 2019 Maarten de Rijke

Theme by Anders NorenUp ↑