Learning user representations based on historical behaviors lies at the core of modern recommender systems. Recent advances in sequential recommenders have convincingly demonstrated high capability in extracting effective user representations from the given behavior sequences. Despite significant progress, we argue that solely modeling the observational behaviors sequences may end up with a brittle and unstable system due to the noisy and sparse nature of user interactions logged. In this paper, we propose to learn accurate and robust user representations, which are required to be less sensitive to (attack on) noisy behaviors and trust more on the indispensable ones, by modeling counterfactual data distribution. Specifically, given an observed behavior sequence, the proposed CauseRec framework identifies dispensable and indispensable concepts at both the fine-grained item level and the abstract interest level. CauseRec conditionally samples user concept sequences from the counterfactual data distributions by replacing dispensable and indispensable concepts within the original concept sequence. With user representations obtained from the synthesized user sequences, CauseRec performs contrastive user representation learning by contrasting the counterfactual with the observational. We conduct extensive experiments on real-world public recommendation benchmarks and justify the effectiveness of CauseRec with multi-aspects model analysis. The results demonstrate that the proposed CauseRec outperforms state-of-the-art sequential recommenders by learning accurate and robust user representations.

CauseRec: Counterfactual User Sequence Synthesis for Sequential Recommendation

CauseRec: Counterfactual User Sequence Synthesis for Sequential Recommendation.

Sequential history of user interactions as well as the context of interactions provide valuable information to recommender systems, for modeling user behavior. Modeling both contexts and sequential information simultaneously, in context-aware sequential recommenders, has been shown to outperform methods that model either one of the two aspects. In long sequential histories, temporal trends are also found within sequences of contexts and temporal gaps that are not modeled by previous methods. In this paper we design new context-aware sequential recommendation methods, based on Stacked Recurrent Neural Networks, that model the dynamics of contexts and temporal gaps. Experiments on two large benchmark datasets demonstrate the advantages of modeling the evolution of contexts and temporal gaps - our models significantly outperform state-of-the-art context-aware sequential recommender systems.

Context-Aware Sequential Recommendations withStacked Recurrent Neural Networks

There is a soaring interest in the news recommendation research scenario due to the information overload. To accurately capture users' interests, we propose to model multi-modal features, in addition to the news titles that are widely used in existing works, for news recommendation. Besides, existing research pays little attention to the click decision-making process in designing multi-modal modeling modules. In this work, inspired by the fact that users make their click decisions mostly based on the visual impression they perceive when browsing news, we propose to capture such visual impression information with visual-semantic modeling for news recommendation. In this paper, we refer to visual impression as the region of the news displayed on the user interface of a news application, which delivers both content and layout information to users. Specifically, we devise the local impression modeling module to simultaneously attend to decomposed details in the impression when understanding the semantic meaning of news title, which could explicitly get close to the process of users reading news. In addition, we inspect the impression from a global view and take structural information, such as the arrangement of different fields and spatial position of different words on the impression, into the modeling of multiple modalities. To accommodate the research of visual impression-aware news recommendation, we extend the text-dominated news recommendation dataset MIND by adding snapshot impression images and will release it to nourish the research field. Extensive comparisons with the state-of-the-art news recommenders along with the in-depth analyses demonstrate the effectiveness of the proposed method and the promising capability of modeling visual impressions for the content-based recommenders.

Why Do We Click: Visual Impression-aware News Recommendation

Effectively representing users lie at the core of modern recommender systems. Since users’ interests naturally exhibit multiple aspects, it is of increasing interest to develop multi-interest frameworks for recommendation, rather than represent each user with an overall embedding. Despite their effectiveness, existing methods solely exploit the encoder (the forward flow) to represent multiple aspects of interests. However, without explicit regularization, the interest embeddings may not be distinct from each other nor semantically reflect representative historical items. Towards this end, we propose the Re4 framework, which leverages the backward flow to reexamine each interest embedding. Specifically, Re4 encapsulates three backward flows, i.e., 1) Re-contrast, which drives each interest embedding to be distinct from other interests using contrastive learning; 2) Re-attend, which ensures the interest-item correlation estimation in the forward flow to be consistent with the criterion used in final recommendation; and 3) Re-construct, which ensures that each interest embedding can semantically reflect the information of representative items that relate to the corresponding interest. We demonstrate the novel forward-backward multi-interest paradigm on ComiRec, and perform extensive experiments on three real-world datasets. Empirical studies validate that Re4 helps to learn learning distinct and effective multi-interest representations.

/pdf/re4-learning-to-re-contrast-re-attend-re-construct-for-multi-2bicjrdu.pdf

Re4: Learning to Re-contrast, Re-attend, Re-construct for Multi-interest Recommendation

In recommender systems, modeling user-item behaviors is essential for user representation learning. Existing sequential recommenders consider the sequential correlations between historically interacted items for capturing users' historical preferences. However, since users' preferences are by nature time-evolving and diversified, solely modeling the historical preference (without being aware of the time-evolving trends of preferences) can be inferior for recommending complementary or fresh items and thus hurt the effectiveness of recommender systems. In this paper, we bridge the gap between the past preference and potential future preference by proposing the future-aware diverse trends (FAT) framework. By future-aware, for each inspected user, we construct the future sequences from other similar users, which comprise of behaviors that happen after the last behavior of the inspected user, based on a proposed neighbor behavior extractor. By diverse trends, supposing the future preferences can be diversified, we propose the diverse trends extractor and the time-aware mechanism to represent the possible trends of preferences for a given user with multiple vectors. We leverage both the representations of historical preference and possible future trends to obtain the final recommendation. The quantitative and qualitative results from relatively extensive experiments on real-world datasets demonstrate the proposed framework not only outperforms the state-of-the-art sequential recommendation methods across various metrics, but also makes complementary and fresh recommendations.

Future-Aware Diverse Trends Framework for Recommendation.

In recommender systems, modeling user-item behaviors is essential for user representation learning. Existing sequential recommenders consider the sequential correlations between historically interacted items for capturing users’ historical preferences. However, since users’ preferences are by nature time-evolving and diversified, solely modeling the historical preference (without being aware of the time-evolving trends of preferences) can be inferior for recommending complementary or fresh items and thus hurt the effectiveness of recommender systems. In this paper, we bridge the gap between the past preference and potential future preference by proposing the future-aware diverse trends (FAT) framework. By future-aware, for each inspected user, we construct the future sequences from other similar users, which comprise of behaviors that happen after the last behavior of the inspected user, based on a proposed neighbor behavior extractor. By diverse trends, supposing the future preferences can be diversified, we propose the diverse trends extractor and the time-aware mechanism to represent the possible trends of preferences for a given user with multiple vectors. We leverage both the representations of historical preference and possible future trends to obtain the final recommendation. The quantitative and qualitative results from relatively extensive experiments on real-world datasets demonstrate the proposed framework not only outperforms the state-of-the-art sequential recommendation methods across various metrics, but also makes complementary and fresh recommendations.

Future-Aware Diverse Trends Framework for Recommendation

Autonomous driving has attracted considerable attention from research and industry communities. Although prototypes of automated vehicles (AVs) are developed, remaining safety issues and functional insufficiencies hinder their market introduction. To obtain reasonably foreseeable scenarios and study human driving policies, many naturalistic driving datasets are proposed. However, no open-source dataset filled with congestion scenarios is publicly available. The paper presents the Aerial Dataset for China's Congested Highways & Expressways (AD4CHE). It contains 5.12 hours of aerial survey data from four different cities in China, with a total driving distance of 6540.7 km. Moreover, overlap and non-overlap cut-in scenarios are distinguished to better describe driver behavior in congestion scenarios. Both types of cut-in scenarios are extracted and parameterized. The Kernel Density Estimator (KDE) is utilized to generate parameter distributions for the scenario-based testing method. Furthermore, the driving behavior in overlap cut-in scenarios is intensively analyzed. The results reveal that the drivers have an evasive maneuver during overlap cut-in of challenging vehicles, and the preferred following distance varies with the relative longitudinal velocity. Both scenario parameterization and driving behavior analysis can contribute to developing and verifying Traffic Jam Pilot (TJP) systems deployed in Chinese traffic situations.

The AD4CHE Dataset and Its Application in Typical Congestion Scenarios of Traffic Jam Pilot Systems

The existing datasets are mostly composed of official documents, statements, news articles, and so forth. So far, only a little attention has been paid to the numerals in financial social comments. Therefore, this paper presents CFinNumAttr, a financial numeral attribute dataset in Chinese via annotating the stock reviews and comments collected from social networking platform. We also conduct several experiments on the CFinNumAttr dataset with state-of-the-art methods to discover the importance of the financial numeral attributes. The experimental results on the CFinNumAttr dataset show that the numeral attributes in social reviews or comments contain rich semantic information, and the numeral clue extraction and attribute classification tasks can make a great improvement in financial text understanding.

Luyao Wang

Papers

Future-Aware Diverse Trends Framework for Recommendation.

Future-Aware Diverse Trends Framework for Recommendation

The AD4CHE Dataset and Its Application in Typical Congestion Scenarios of Traffic Jam Pilot Systems

A Chinese Fine-grained Financial Event Extraction Dataset