The importance of incorporating interpretability into machine learning models has been increasingly emphasized. While previous literature has typically focused on feature level interpretability, such as analyzing which features are important and how they influence the final decision, real-world applications often require domain level interpretability, which relates to a group of features. Domain-level interpretability holds the potential for enhanced informativeness and comprehensibility. Unfortunately, there has been limited research in this direction. In this paper, we address this issue and introduce our proposed method DIDE, which obtains domain-level interpretability from domain-specific latent embeddings. To enhance the effectiveness of the framework, we draw inspiration from the gradient smooth philosophy and propose noisy injection in the embedding space, resulting in smoothed interpretability. We conduct extensive experiments to validate the effectiveness of DIDE, and demonstrate its applications in assisting daily business tasks in Alipay.

Though neural networks have achieved impressive prediction performance, it's still hard for people to understand what neural networks have learned from the data. The black-box property of neural networks already becomes one of the main obstacles preventing from being applied to many high-stakes applications, such as finance and medicine that have critical requirement on the model transparency and interpretability. In order to enhance the explainability of neural networks, we propose a neural rule learning method-Feature Interactive Neural Rule Learning (FINRule) to incorporate the expressivity of neural networks and the interpretability of rule-based systems. Specifically, we conduct rule learning as differential discrete combination encoded by a feedforward neural network, in which each layer acts as a logical operator of explainable decision conditions. The first hidden layer can act as sharable atomic conditions which are connected to next hidden layer for formulating decision rules. Moreover, we propose to represent both atomic condition and rules with contextual embeddings, with aim to enrich the expressivity power by capturing high-order feature interactions. We conduct comprehensive experiments on real-world datasets to validate both effectiveness and explainability of the proposed method.

Decision rules have been widely applied in industrial applications such as finance, medicine, and biology, due to the critical requirement of interpretability. In order to make decision rules easier and more widely used in financial scenarios, an automatic intelligent rule system with rule learning and rule management capabilities is needed. However, the rule system for financial applications has distinctive challenges both in algorithms and systems. From the algorithm perspective, due to the characteristics of the financial data and scenarios, the rule learning algorithm faces the class-imbalanced issue, the scalability issue, and the diversity of optimization objectives. From the system perspective, a flexible rule learning and management framework is needed to adapt to fast-changing financial applications with heterogenous data, and engineering optimization is required to ensure the time and space efficiency of rule learning. In this work, we focus on developing a Rule-based Decision System (RDS) to deal with the algorithmic and systematic challenges mentioned above. RDS covers the full life cycle of the decision rules, including the rule learning module, rule management module, and rule deployment module. Moreover, the rule system offers an interactive interface to allow users to integrate the expert experiences into the decision rules and realize the human-in-the-loop. The RDS has been deployed on one of the world’s largest trading and money transfer platforms, serving hundreds of millions of users and transactions.

In recommendation scenarios, there are two long-standing challenges, i.e., selection bias and data sparsity, which lead to a significant drop in prediction accuracy for both Click-Through Rate (CTR) and post-click Conversion Rate (CVR) tasks. To cope with these issues, existing works emphasize on leveraging Multi-Task Learning (MTL) frameworks (Category 1) or causal debiasing frameworks (Category 2) to incorporate more auxiliary data in the entire exposure/inference space $\mathcal{D}$ or debias the selection bias in the click/training space ${\mathcal{O}}$. However, these two kinds of solutions cannot effectively address the not-missing-at-random problem and debias the selection bias in ${\mathcal{O}}$ to fit the inference in $\mathcal{D}$. To fill the research gaps, we propose a Direct entire-space Causal Multi-Task framework, namely DCMT, for post-click conversion prediction in this paper. Specifically, inspired by users’ decision process of conversion, we propose a new counterfactual mechanism to debias the selection bias in $\mathcal{D}$, which can predict the factual CVR and the counterfactual CVR under the soft constraint of a counterfactual prior knowledge. Extensive experiments demonstrate that our DCMT can improve the state-of-the-art methods by an average of 1.07% in term of CVR AUC on the offline datasets and 0.75% in term of PV-CVR on the online A/B test (the Alipay Search). Such improvements can increase millions of conversions per week in real industrial applications, e.g., the Alipay Search.

The cold-start problem has been commonly recognized in recommendation systems and studied by following a general idea to leverage the abundant interaction records of warm users to infer the preference of cold users. However, the performance of these solutions is limited by the amount of records available from warm users to use. Thus, building a recommendation system based on few interaction records from a few users still remains a challenging problem for unpopular or early-stage recommendation platforms. This paper focuses on solving the few-shot recommendation problem for news recommendation based on two observations. First, news at different platforms (even in different languages) may share similar topics. Second, the user preference over these topics is transferable across different platforms. Therefore, we propose to solve the few-shot news recommendation problem by transferring the user-news preference from a many-shot source domain to a few-shot target domain. To bridge two domains that are even in different languages and without any overlapping users and news, we propose a novel unsupervised cross-lingual transfer model as the news encoder that aligns semantically similar news in two domains. A user encoder is constructed on top of the aligned news encoding and transfers the user preference from the source to target domain. Experimental results on two real-world news recommendation datasets show the superior performance of our proposed method on addressing few-shot news recommendation, comparing to the baselines. The source code can be found at https://github.com/taichengguo/Few-shot-NewsRec.

Entity synonym discovery (ESD) from text corpus is an essential problem in many entity-leveraging applications, e.g., web search and question answering. This paper aims to address three limitations that widely exist in the current ESD solutions: 1) the lack of effective utilization for synonym set information; 2) the feature extraction of entities from restricted receptive fields; and 3) the incapacity to capture higher-order contextual information. We propose a novel set-aware ESD model that enables a flexible receptive field for ESD by making a breakthrough in using entity synonym set information. The contextual information of entities and entity synonym sets are arranged by a two-level network from which entities and entity synonym sets can be mapped into the same embedding space to facilitate ESD by encoding the high-order contexts from flexible receptive fields. Extensive experimental results on public datasets show that our model consistently outperforms the state-of-the-art with significant improvement.

Pairwise ranking models have been widely used to address recommendation problems. The basic idea is to learn the rank of users' preferred items through separating items into positive samples if user-item interactions exist, and negative samples otherwise. Due to the limited number of observable interactions, pairwise ranking models face serious class-imbalance issues. Our theoretical analysis shows that current sampling-based methods cause the vertex-level imbalance problem, which makes the norm of learned item embeddings towards infinite after a certain training iterations, and consequently results in vanishing gradient and affects the model inference results. We thus propose an efficient Vital Negative Sampler (VINS) to alleviate the class-imbalance issue for pairwise ranking model, in particular for deep learning models optimized by gradient methods. The core of VINS is a bias sampler with reject probability that will tend to accept a negative candidate with a larger degree weight than the given positive item. Evaluation results on several real datasets demonstrate that the proposed sampling method speeds up the training procedure 30% to 50% for ranking models ranging from shallow to deep, while maintaining and even improving the quality of ranking results in top-N item recommendations.

Decision rules have been successfully used in various classification applications because of their interpretability and efficiency. In many real-world scenarios, especially in industrial applications, it is necessary to generate rule sets under certain constraints, such as confidence constraints. However, most previous rule mining methods only emphasize the accuracy of the rule set but take no consideration of these constraints. In this paper, we propose a Confidence-constraint Rule Set Learning (CRSL) framework consisting of three main components, i.e. rule miner, rule ranker, and rule subset selector. Our method not only considers the trade-off between confidence and coverage of the rule set but also considers the trade-off between interpretability and performance. Experiments on benchmark data and large-scale industrial data demonstrate that the proposed method is able to achieve better performance (6.7% and 8.8% improvements) and competitive interpretability when compared with other rule set learning methods.

Machine learning methods for fraud detection have achieved impressive prediction performance, but often sacrifice critical interpretability in many applications. In this work, we propose to learn interpretable models for fraud detection as a simple rule set. More specifically, we design a novel neural rule learning method by building a condition graph with an expectation to capture the high-order feature interactions. Each path in this condition graph can be regarded as a single rule. Inspired by the key idea of meta learning, we combine the neural rules with rules extracted from the tree-based models in order to provide generalizable rule candidates. Finally, we propose a flexible rule set learning framework by designing a greedy optimization method towards maximizing the recall number of fraud samples with a predefined criterion as the cost. We conduct comprehensive experiments on large-scale industrial datasets. Interestingly, we find that the neural rules and rules extracted from tree-based models can be complementary to each other to improve the prediction performance.

Entity synonym discovery (ESD) from text corpus is an essential problem in many entity-leveraging applications. This paper aims to address three limitations that widely exist in the current ESD solutions: 1) the lack of effective utilization for synonym set information; 2) the feature extraction of entities from restricted receptive fields; and 3) the incapacity to capture higher-order contextual information. We propose a novel set-aware ESD model that enables a flexible receptive field for ESD by using entity synonym set information and constructing a two-level network. Extensive experimental results on public datasets show that our model consistently outperforms the state-of-the-art with significant improvement.

Recent years have witnessed increasing attention on the application of graph alignment to on-Web tasks, such as knowledge graph integration and social network linking. Despite achieving remarkable performance, prevailing graph alignment models still suffer from noisy supervision, yet how to mitigate the impact of noise in labeled data is still under-explored. The negative sampling based noise discrimination model has been a feasible solution to detect the noisy data and filter them out. However, due to its sensitivity to the sampling distribution, the negative sampling based noise discrimination model would lead to an inaccurate decision boundary. Furthermore, it is difficult to find an abiding threshold to separate the potential positive (benign) and negative (noisy) data in the whole training process. To address these important issues, in this paper, we design a non-sampling discrimination model resorting to the unbiased risk estimation of positive-unlabeled learning to circumvent the harmful impact of negative sampling. We also propose to select the appropriate potential positive data at different training stages by an adaptive filtration threshold enabled by curriculum learning, for maximally improving the performance of alignment model and non-sampling discrimination model. Extensive experiments conducted on several real-world datasets validate the effectiveness of our proposed method.

Web personalization, e.g., recommendation or relevance search, tailoring a service/product to accommodate specific online users, is becoming increasingly important. Inductive personalization aims to infer the relations between existing entities and unseen new ones, e.g., searching relevant authors for new papers or recommending new items to users. This problem, however, is challenging since most of recent studies focus on transductive problem for existing entities. In addition, despite some inductive learning approaches have been introduced recently, their performance is sub-optimal due to relatively simple and inflexible architectures for aggregating entity’s content. To this end, we propose the inductive contextual personalization (ICP) framework through contextual relation learning. Specifically, we first formulate the pairwise relations between entities with a ranking optimization scheme that employs neural aggregator to fuse entity’s heterogeneous contents. Next, we introduce a node embedding term to capture entity’s contextual relations, as a smoothness constraint over the prior ranking objective. Finally, the gradient descent procedure with adaptive negative sampling is employed to learn the model parameters. The learned model is capable of inferring the relations between existing entities and inductive ones. Thorough experiments demonstrate that ICP outperforms numerous baseline methods for two different applications, i.e., relevant author search and new item recommendation.

Cross-lingual entity alignment aims at associating semantically similar entities in knowledge graphs with different languages. It has been an essential research problem for knowledge integration and knowledge graph connection, and been studied with supervised or semi-supervised machine learning methods with the assumption of clean labeled data. However, labels from human annotations often include errors, which can largely affect the alignment results. We thus aim to formulate and explore the robust entity alignment problem, which is non-trivial, due to the deficiency of noisy labels. Our proposed method named REA (Robust Entity Alignment) consists of two components: noise detection and noise-aware entity alignment. The noise detection is designed by following the adversarial training principle. The noise-aware entity alignment is devised by leveraging graph neural network based knowledge graph encoder as the core. In order to mutually boost the performance of the two components, we propose a unified reinforced training strategy to combine them. To evaluate our REA method, we conduct extensive experiments on several real-world datasets. The experimental results demonstrate the effectiveness of our proposed method and also show that our model consistently outperforms the state-of-the-art methods with significant improvement on alignment accuracy in the noise-involved scenario.

This paper targets on designing a query-based dataset recommendation system, which accepts a query denoting a user's research interest as a set of research papers and returns a list of recommended datasets that are ranked by the potential usefulness for the user's research need. The motivation of building such a system is to save users from spending time on heavy literature review work to find usable datasets.We start by constructing a two-layer network: one layer of citation network, and the other layer of datasets, connected to the firstlayer papers in which they were used. A query highlights a set of papers in the citation layer. However, answering the query as a naive retrieval of datasets linked with these highlighted papers excludes other semantically relevant datasets, which widely exist several hops away from the queried papers. We propose to learn representations of research papers and datasets in the two-layer network using heterogeneous variational graph autoencoder, and then compute the relevance of the query to the dataset candidates based on the learned representations. Our ranked datasets shown in extensive evaluation results are validated to be more truly relevant than those obtained by naive retrieval methods and adoptions of existing related solutions.

Entity alignment associates entities in different knowledge graphs if they are semantically same, and has been successfully used in the knowledge graph construction and connection. Most of the recent solutions for entity alignment are based on knowledge graph embedding, which maps knowledge entities in a low-dimension space where entities are connected with the guidance of prior aligned entity pairs. The study in this paper focuses on two important issues that limit the accuracy of current entity alignment solutions: 1) labeled data of priorly aligned entity pairs are difficult and expensive to acquire, whereas abundant of unlabeled data are not used; and 2) knowledge graph embedding is affected by entity's degree difference, which brings challenges to align high frequent and low frequent entities. We propose a semi-supervised entity alignment method (SEA) to leverage both labeled entities and the abundant unlabeled entity information for the alignment. Furthermore, we improve the knowledge graph embedding with awareness of the degree difference by performing the adversarial training. To evaluate our proposed model, we conduct extensive experiments on real-world datasets. The experimental results show that our model consistently outperforms the state-of-the-art methods with significant improvement on alignment accuracy.

With the advances in technology and smart devices, more and more attention has been paid to model spatial correlations, temporal dynamics, and friendship influence over point-of-interest (POI) checkins. Besides directly capturing general user's checkin behavior, existing works mostly highlight the intrinsic feature of POIs, i.e., spatial and temporal dependency. Among them, the family of methods based on Markov chain can capture the instance-level interaction between a pair of POI checkins, while recurrent neural network (RNN) based approaches (state-of-the-art) can deal with flexible length of checkin sequence. However, the former is not good at capturing high-order POI transition dependency, and the latter cannot distinguish the exact contribution of each POI in a historical checkin sequence. Moreover, in recurrent neural networks, local and global information is propagated along the sequence through one bottleneck i.e., hidden states only.In this work, we design a novel model to enforce contextual constraints on sequential data by designing a spatial and temporal attention mechanisms over recurrent neural network that leverages the importance of POIs visited by users in given time interval and geographical distance in successive checkins. Attention mechanism helps us to learn which POIs bounded by time difference and spatial distance in user checkin history are important for the prediction of next POI. Moreover, we also consider periodicity and friendship influence in our model design. Experimental results on two real location based social networks Gowalla, and BrightKite show that our proposed method outperforms the existing state-of-the-art deep neural network methods for next POI prediction and understanding user transition behavior. We also analyze the sensitivity of parameters including context window for capturing sequential effect, temporal context window for estimating temporal attention and spatial context window for estimating spatial attention respectively.

In this paper, we study the problem of author identification in big scholarly data, which is to effectively rank potential authors for each anonymous paper by using historical data. Most of the existing de-anonymization approaches predict relevance score of paper-author pair via feature engineering, which is not only time and storage consuming, but also introduces irrelevant and redundant features or miss important attributes. Representation learning can automate the feature generation process by learning node embeddings in academic network to infer the correlation of paper-author pair. However, the learned embeddings are often for general purpose (independent of the specific task), or based on network structure only (without considering the node content). To address these issues and make a further progress in solving the author identification problem, we propose Camel, a content-aware and meta-path augmented metric learning model. Specifically, first, the directly correlated paper-author pairs are modeled based on distance metric learning by introducing a push loss function. Next, the paper content embedding encoded by the gated recurrent neural network is integrated into the distance loss. Moreover, the historical bibliographic data of papers is utilized to construct an academic heterogeneous network, wherein a meta-path guided walk integrative learning module based on the task-dependent and content-aware Skipgram model is designed to formulate the correlations between each paper and its indirect author neighbors, and further augments the model. Extensive experiments demonstrate that Camel outperforms the state-of-the-art baselines. It achieves an average improvement of 6.3% over the best baseline method.

Data sparsity and cold-start problems are prevalent in recommender systems. To address such problems, both the observable explicit social information (e.g., user-user trust connections) and the inferable implicit correlations (e.g., implicit neighbors computed by similarity measurement) have been introduced to complement user-item ratings data for improving the performances of traditional model-based recommendation algorithms such as matrix factorization. Although effective, (1) the utilization of the explicit user-user social relationships suffers from the weakness of unavailability in real systems such as Netflix or the issue of sparse observable content like 0.03% trust density in Epinions, thus there is no or little explicit social information that can be employed to improve baseline model in real applications; (2) the current similarity measurement approaches focus on inferring implicit correlations between a user (item) and their direct neighbors or top-k similar neighbors based on user-item ratings bipartite network, so that they fail to comprehensively unfold the indirect potential relationships among users and items. To solve these issues regarding both explicit/implicit social recommendation algorithms, we design a joint model of matrix factorization and implicit walk integrative learning, i.e., ImWalkMF, which only uses explicit ratings information yet models both direct rating feedbacks and multiple direct/indirect implicit correlations among users and items from a random walk perspective. We further propose a combined strategy for training two independent components in the proposed model based on sampling. The experimental results on two real-world sparse datasets demonstrate that ImWalkMF outperforms the traditional regularized/probabilistic matrix factorization models as well as other competitive baselines that utilize explicit/implicit social information.

To address the issue of data sparsity and cold-start in recommender system, social information (e.g., user-user trust links) has been introduced to complement rating data for improving the performances of traditional model-based recommendation techniques such as matrix factorization (MF) and Bayesian personalized ranking (BPR). Although effective, the utilization of the explicit user-user relationships extracted directly from such social information has three main limitations. First, it is difficult to obtain explicit and reliable social links. Only a small portion of users indicate explicitly their trusted friends in recommender systems. Second, the “cold-start” users are “cold” not only on rating but also on socializing. There is no significant amount of explicit social information that can be useful for “cold-start” users. Third, an active user can be socially connected with others who have different taste/preference. Direct usage of explicit social links may mislead recommendation. To address these issues, we propose to extract implicit and reliable social information from user feedbacks and identify top-k semantic friends for each user. We incorporate the top-k semantic friends information into MF and BPR frameworks to solve the problems of ratings prediction and items ranking, respectively. The experimental results on three real-world datasets show that our proposed approaches achieve better results than the state-of-the-art MF with explicit social links (with 3.0% improvement on RMSE), and social BPR (with 9.1% improvement on AUC).