SIGIR Workshop on eCommerce  will take place on Thursday, July 30, 2020

SIGIR eCom'20 Workshop Program

Asia Session (July 30, All times below in GMT +8)                 Session Chairs - Tracy King & Weihua Luo & Shervin Malmasi
8:55 am CST
8:55 pm EDT
5:55 pm PDT
Opening Remarks
9:05 am
Natural Language Processing R&D for E-commerce
Luo Si    Alibaba, USA

Abstract: Natural Language Processing (NLP) and related technologies are critical for the success of Internet business like e-commerce. Alibaba’s NLP R&D aims at supporting the business demands of Alibaba’s eco-system, creating new opportunities for Alibaba’s partners and advancing the state-of-the-art of NLP technologies. This talk will introduce our efforts to build NLP technique platform and machine translation (MT) platform that power Alibaba’s eco-system. Furthermore, some recent research work will be presented on product title compression with user-log information, sentiment classification with questions & answers, machine reading comprehension in real-world custom service, and cascade ranking for large-scale e-commerce search. The R&D work attracts hundreds of millions of users and generates significant business value every day.
9:50 am Contributed Talk 1
Counterfactual Learning to Rank using Heterogeneous Treatment Effect Estimation
Mucun Tian, Chun Guo, Vito Ostuni and Zhen Zhu
10:10 am Contributed Talk 2
Query Transformation for Multi-Lingual Product Search
Qie Hu, Hsiang-Fu Yu, Vishnu Narayanan, Ivan Davchev, Rahul Bhagat and Inderjit Dhillon
10:30 am Coffee Break
11:00 am Keynote
Challenges in Multimodal Conversational Search
Tat-Seng Chua    National University of Singapore, Singapore

Abstract: Information search has been evolving from mostly unidirectional and text-based to interactive and multimodal. Recently, there is also a growing interest in all matters conversational. By incorporating multimodal conversation, it offers users a natural way to query the system by combining text/speech, images/videos and possibly gesture. It also helps to tackle the basic asymmetric problem in search by injecting conversation to resolve ambiguities in search and recommendation. However, the evolution from traditional IR to conversational IR faces many challenges. Among them are the need to develop new models and framework to: model conversational context and history, integrate domain knowledge and user models; conduct interactive IR and QA, develop intervention strategy to incorporate conversation into browsing; and integrate conversation with recommendation, database search and Web search. Moreover, there are the issues of resources, methodologies and biasness in evaluating (multi-turn) conversational search systems. This talk presents current research and challenges in tackling these problems with pointers towards future research.
11:45 am Invited Talk
The challenges of Search and Recommendation in Alibaba International Business
Anxiang Zeng    Alibaba, China

Abstract: The international search team of Alibaba supports the scenarios of search and recommendation of the international business branches of Alibaba, such as AliExpliss, Lazada. In these scenarios, the team faces the challenges different from Taobao, which is a local marketplace. The biggest one is relevance, which is the basic component of search. However, due to the lack of understanding of language and intuitive experience, it is difficult to solve it by traditional methods. The team tries to leverage knowledge Graph and other methods to reduce the workload of labelling. In addition, there are a large number of users with sparse behavior in these scenarios, and the performance of the algorithm will benefit from better understanding and modeling them. The team has made preliminary attempts for different aspects of related challenges, and achieved good results on some issues. In this talk, I will introduce the main framework of the search and recommendation algorithm applied in the international business of Alibaba, and presents our innovative work for the new challenges.
12:20 pm Coffee Break
12:40 pm Data Challenge Papers
A Multi-Modal Late Fusion Model for E-Commerce Product Classification and Retrieval
Shuo Wang, Ye Bi and Zhongrui Fan

Large Scale Multimodal Classification Using an Ensemble of Transformer Models and Co-Attention
Varnith Chordia and Vijay Kumar
1:10 pm - 2:10 pm Virtual Discussion Session (60 minutes)
   9. Bias Correction for Supervised Learning in Email Marketing  [PDF]
        Moumita Sinha, Yancheng Li, Wei Shung Chung and Paul Hsiung (Adobe)

 10. Atlas: A Dataset and Benchmark for E-commerce Clothing Product Categorization  [PDF]
         Venkatesh Umaashankar (Ericsson Research), Girish Shanmugam S (Uppsala Uni.) and Aditi Prakash (Uni. of Colorado, Boulder)

 11. Fantastic Embeddings and How to Align Them: Zero-Shot Inference in a Multi-Shop Scenario  [PDF]
         Federico Bianchi, Jacopo Tagliabue, Bingqing Yu, Luca Bigon and Ciro Greco (Coveo Labs)

 12. Revenue, Relevance, Arbitrage & More: Joint Optimization Framework for Search Experiences in Two-Sided Marketplaces  [PDF]
         Andrew Stanton (Etsy Inc.), Akhila Ananthram (Etsy Inc.), Congzhe Su (Etsy Inc.) and Liangjie Hong (LinkedIn)

 13. Exploiting Neural Query Translation into Cross Lingual Information Retrieval  [PDF]
         Liang Yao, Baosong Yang, Haibo Zhang, Weihua Luo and Boxing Chen (Alibaba)

 17. Online Product Feature Recommendations with Interpretable Machine Learning  [PDF]
         Mingming Guo, Nian Yan, Xiquan Cui, Simon Hughes and Khalifeh Al Jadda (Home Depot)

 19. Constraint Translation Candidates: A Bridge between Neural Query Translation and Cross-lingual Information Retrieval  [PDF]
         Tianchi Bi, Liang Yao, Baosong Yang, Haibo Zhang, Weihua Luo and Boxing Chen (Alibaba)

   7. Discriminative Pre-training for Low Resource Title Compression in Conversational Grocery  [PDF]
        Snehasish Mukherjee, Phaniram Sayapaneni and Shankar Subramanya (Walmart Labs)

Americas/ Europe Session (July 30, All times below in EDT - GMT-4)                 Session Chairs - Dietmar Jannach & Surya Kallumadi & Tracy King
9:00 am EDT
6:00 am PDT
3:00 pm CET
Student Papers Spotlight Session

Improved Session based Recommendation using Graph-based Item Embedding
Madiraju Srilakshmi, Gourab Chowdhury and Sudeshna Sarkar

Aspect Category Detection in Product Reviews using Contextual Representation
Shiva Ramezani, Razieh Rahimi and James Allan

A Comparison of Supervised Learning to Match Methods for Product Search
Fatemeh Sarvi, Nikos Voskarides, Lois Mooiman, Sebastian Schelter and Maarten de Rijke

Analyzing and Predicting Purchase Intent in E-commerce: Anonymous vs. Identified Customers
Mariya Hendriksen, Ernst Kuiper, Pim Nauts, Sebastian Schelter and Maarten de Rijke
9:40 am Coffee Break
10:00 am EDT
07:00 am PDT
04:00 pm CET
Report on SIGIR eCom’20 Rakuten Data Challenge

Data Challenge Papers
Synerise at SIGIR Rakuten Data Challenge 2020: Efficient Manifold Density Estimator for Cross-Modal Retrieval
Barbara Rychalska and Jacek Dąbrowski

CBB-FE, CamemBERT and BiT Feature Extraction for Multimodal Product Classification and Retrieval
Hou Wei Chou, Younghun Lee, Lei Chen, Yandi Xia and Wei Te Chen

Deep Multi-level Fusion Learning Framework for Multi-modal Product Classification
Ekansh Verma, Souradip Chakraborty and Vinodh Motupalli
11:00 am Invited Talk
Shaping Recommendation on Multi-stakeholder Platforms
Rishabh Mehrotra    Spotify, UK

Abstract: Multi-sided marketplaces facilitate efficient interactions between multiple stakeholders, including e.g. buyers and retailers (Amazon), guests and hosts (AirBnb), riders and drivers (Uber), and listeners and artists (Spotify). Recommender systems powering online multi-stakeholder platforms often face the challenge of jointly optimizing multiple objectives, in an attempt to efficiently match suppliers and consumers. In this talk, I will touch upon some recent advances in designing recommendation systems that power such platforms, including (1) multi-objective bandits, (2) user & content aware reward modeling, (3) role of consumption & supplier diversity and (4) interplay between stakeholder objectives. I will discuss insights from large scale experiments, and highlight important research directions.
11:40 am Coffee Break
11:55 am Opening Remarks
12:05 pm
Beyond Being Accurate: Solving Real-World Recommendation Problems with Neural Modeling
Ed Chi    Google, USA

Abstract: Fundamental improvements in recommendation and search ranking have been much harder to come by, when compared with progress on other long-standing AI problems such as visual/audio machine perception and machine translation. Some reasons include: (1) large amounts of data making training difficult, yet having (2) power-law shaped, noisy, and sparse labels; (3) changing dynamics of context such as user preferences and items; (4) optimizing for multiple objectives, and (5) low-latency requirement for a recommendation response. Beyond that, one huge challenge is devising approaches for (6) more inclusive and fairer models to multiple stakeholders. How do we make progress?
While search and recommendation engines are increasingly more intelligent, the above challenges are evolving the two technology stacks toward each other, and there is increasingly a blurring of the boundary between these two approaches to information seeking. This blurring has resulted in critical re-thinking in how to architect the systems by merging and sharing neural modeling techniques common to both types of systems.
In this talk, I will touch upon many recent advances in neural modeling techniques for recommendations and their impact in Google products covering ~250 improvements over the last 3 years, including:
- policy gradient RL techniques with off-policy correction in recurrent recommendation models;
- multi-task models with gated mixture of experts;
- diversification and slate optimization with determinantal point processes;
- large output item spaces with Neural Deep Retrieval;
- utilizing TPUs for large sparse models;
- adversarial approaches for ML Fairness for Classifiers and Recommenders.

12:50 pm Panel Discussion
When Search meets Recommendations
Vanessa Murdock     Amazon, USA
Maarten de Rijke     University of Amsterdam & Ahold Delhaize, The Netherlands
Tracy King     Adobe, USA
Joe Konstan     University of Minnesota, USA
1:40 pm Coffee Break
2:00 pm
Inclusive Search and Recommendations
Nadia Fawaz    Pinterest, USA

Abstract: Machine learning powers many advanced search and recommendation systems, and user experience strongly depends on how well ML systems perform across all data segments. This performance can be impacted by biases, which can lead to a subpar experience for subsets of users, content providers, applications or use cases. Biases may arise at different stages in machine learning systems, from existing societal biases in the data, to biases introduced by the data collection or modeling processes. These biases may impact the performance of various components of ML systems, from offline training, to evaluation and online serving in production systems. Specific techniques have been developed to help reduce bias at each stage of an ML system. We will describe sources of bias in ML technology, why addressing bias matters, and techniques to mitigate bias, with examples from our work on inclusive AI at Pinterest. Mitigating bias in machine learning systems is crucial to successfully achieve our mission to "bring everyone the inspiration to create a life they love".
2:45 pm Contributed Talk 3
Context-Aware Learning to Rank with Self-Attention
Przemysław Pobrotyn, Tomasz Bartczak, Mikołaj Synowiec, Radosław Białobrzeski and Jarosław Bojar
3:05 pm Contributed Talk 4
Light Feed-Forward Networks for Shard Selection in Large-scale Product Search
Heran Lin, Pengcheng Xiong, Danqing Zhang, Fan Yang, Ryoichi Kato, Mukul Kumar, William Headden and Bing Yin
3:25am Coffee Break
3:45 pm Contributed Talk 5
Shopping in the Multiverse: A Counterfactual Approach to In-Session Attribution
Jacopo Tagliabue and Bingqing Yu
4:05 pm Invited Talk
End-to-end machine learning systems for e-commerce
Keld Lundgaard    Salesforce, USA

Abstract: Machine learning has been a part of e-commerce from the beginning, through search (NLP), recommendation systems (collaborative filtering), market basket analysis, optimal routing, and many other parts of the e-commerce system. However, with the advent of deep learning, a number of new parts of the merchant experience are becoming augmented by machine intelligence, with attribute models, complete the set generation, and catalog management. With these new parts, it is time to look at the e-commerce machine learning perspective from a holistic perspective and ask, how we can connect all these systems such that each part can learn from each other in an end-to-end fashion.
4:35 pm Invited Talk
A unified success and experimentation framework
Pranam Kolari    Walmart labs, USA

Abstract: The concepts of North star, success metrics, machine learning objective, and experimentation are common to most industrial setups. In this talk, we bring these concepts together through a simple unified framework, layout how they impact the product experience they are aimed to measure and improve, and motivate the need for more contributions to the area, both from academia and industrial research.
5:05 pm - 5:30 pm Group Discussion and Closing