The field of graph foundation models (GFMs) has seen a dramatic rise in interest in recent years. Their powerful generalization ability is believed to be endowed by self-supervised pre-training and downstream tuning techniques. There is a wide variety of knowledge patterns embedded in the graph data, such as node properties and clusters, which are crucial for learning generalized representations for GFMs. We present a comprehensive survey of self-supervised GFMs from a novel knowledge-based perspective. Our main contribution is a knowledge-based taxonomy that categorizes self-supervised graph models by the specific graph knowledge utilized: microscopic (nodes, links, etc.), mesoscopic (context, clusters, etc.), and macroscopic (global structure, manifolds, etc.). It covers a total of 9 knowledge categories and 300 references for self-supervised pre-training as well as various downstream tuning strategies. Such a knowledge-based taxonomy allows us to more clearly re-examine potential GFM architectures, including large language models (LLMs), as well as provide deeper insights for constructing future GFMs.

In recommender systems, the patterns of user behaviors (e.g., purchase, click) may vary greatly in different contexts (e.g., time and location). This is because user behavior is jointly determined by two types of factors: intrinsic factors , which reflect consistent user preference, and extrinsic factors , which reflect external incentives that may vary in different contexts. Differentiating between intrinsic and extrinsic factors helps learn user behaviors better. However, existing studies have only considered differentiating them from a single, pre-defined context (e.g., time or location), ignoring the fact that a user’s extrinsic factors may be influenced by the interplay of various contexts at the same time. In this article, we propose the intrinsic-extrinsic disentangled recommendation (IEDR) model, a generic framework that differentiates intrinsic from extrinsic factors considering various contexts simultaneously, enabling more accurate differentiation of factors and hence the improvement of recommendation accuracy. IEDR contains a context-invariant contrastive learning component to capture intrinsic factors, and a disentanglement component to extract extrinsic factors under the interplay of various contexts. The two components work together to achieve effective factor learning. Extensive experiments on real-world datasets demonstrate IEDR’s effectiveness in learning disentangled factors and significantly improving recommendation accuracy by up to 4% in NDCG.

Emotion Cause Pair Extraction (ECPE) aims to extract emotions and their causes from a document. Powerful emotion and cause extraction abilities have proven essential in achieving accurate ECPE. However, most existing methods employ shared feature learning of emotion extraction and cause extraction, which can harm the abilities of both tasks as they focus on different information (i.e., task-specific features). Moreover, shared feature learning of the two tasks also leads to the label imbalance problem. To address these issues, this paper proposes a multi-task learning framework named Hierarchical Shared Encoder with Task-specific Transformer Layer Selection (HSE-TTLS). The model achieves ECPE via two subtasks: Emotion Extraction (EE) and Emotion Cause Extraction (ECE). The design of two subtasks for ECPE corresponds to the fact that cause clauses are emotion-dependent and significantly alleviates the label imbalance problem. To effectively extract task-specific features for EE and ECE, we employ BERT as the token-level encoder and select task-specific optimal layers for the two subtasks. Focal loss is used as the objective function for EE to further alleviate the label imbalance problem. Extensive experiments on benchmark ECPE corpus demonstrate the effectiveness of HSE-TTLS, which outperforms state-of-the-art baseline methods by at least 1.56% on the F1 score.

The task of entity alignment between knowledge graphs (KGs) aims to identify every pair of entities from two different KGs that represent the same entity. Many machine learning-based methods have been proposed for this task. However, to our best knowledge, existing methods all requiremanually craftedseed alignments, which are expensive to obtain. In this paper, we propose the first fully automatic alignment method named AutoAlign, which does not require any manually crafted seed alignments. Specifically, for predicate embeddings, AutoAlign constructs a predicate-proximity-graph with the help of large language models to automatically capture the similarity between predicates across two KGs. For entity embeddings, AutoAlign first computes the entity embeddings of each KG independently using TransE, and then shifts the two KGs' entity embeddings into the same vector space by computing the similarity between entities based on their attributes. Thus, both predicate alignment and entity alignment can be done without manually crafted seed alignments. AutoAlign is not only fully automatic, but also highly effective. Experiments using real-world KGs show that AutoAlign improves the performance of entity alignment significantly compared to state-of-the-art methods. Our source code is available at ruizhang-ai/AutoAlign.

Region similarity learning plays an essential role in applications such as business site selection, region recommendation, and urban planning. Earlier studies mainly represent regions as bags of points of interest (POIs) for region similarity comparisons, which cannot fully exploit the spatial features of the regions. Recently, researchers propose to use deep neural networks to exploit spatial features such as POI geo-coordinates and categories, which have produced more accurate and robust region similarity learning results. However, many useful features such as the height and size of a POI, and the distance and relative importance between the POIs, are still overlooked in these methods. To take advantage of such features, we propose to represent regions as graphs, where nodes are POIs with rich features such as height, size, and hexagonal coordinates, while edges are the relationships between POIs formulated by their road network distances. To capture POIs’ importance, we weigh them by their height and size. Since there is limited availability of ground-truth region similarity data, we propose a contrastive learning-based multi-relational graph neural network (C-MPGCN) for region similarity learning based on the graph representations. To generate data for model training, we propose a soft graph edit distance (SGED) based algorithm to generate triples of similar and dissimilar graphs of a given graph (representing a given region) based on the POI weights. Experimental results show that C-MPGCN outperforms the state-of-the-art methods for region similarity learning consistently with an improvement of at least 8.6% and 9.4% in terms of MRR and HR@1, respectively.

Sequential recommendation systems predict the next item that a user will interact with according to the sequential patterns inferred from historical interacted items. Benefiting from the self-attention mechanism, BERT has achieved success in sequential recommendation. Nevertheless, it is still challenging to leverage multiple types of information (e.g., item IDs, attributes, timestamps) effectively under the BERT framework. Existing methods fuse different types of information as the input of BERT by directly adding or concatenating their embedding vectors. This way of fusion ignores the differences of different types of information in sequential pattern inference, which can cause the mutual interference problem and hinder the effective utilization of multi-information. In this work, we propose a Dual Disentangled Attention (DDA) based BERT model, called DDA-BERT, for better leveraging multi-information. Our model performs disentangled information modeling from two aspects: 1) different views to describe items using content information (e.g., using IDs or using attributes) are modeled in different attention heads, respectively; 2) different sequential information (time intervals, relative position distances) modeling is disentangled from content information modeling while calculating the attention scores. As a result, DDA-BERT can effectively prevent the mutual interference problem to improve the prediction accuracy. Extensive experiments on three benchmark datasets show that DDA-BERT consistently outperforms the state-of-the-art baselines by up to 30%.

Detecting beneficial feature interactions is essential in recommender systems, and existing approaches achieve this by examining all the possible feature interactions. However, the cost of examining all the possible higher-order feature interactions is prohibitive (exponentially growing with the order increasing). Hence existing approaches only detect limited order (e.g., combinations of up to four features) beneficial feature interactions, which may miss beneficial feature interactions with orders higher than the limitation. In this paper, we propose a hypergraph neural network based model named HIRS. HIRS is the first work that directly generates beneficial feature interactions of arbitrary orders and makes recommendation predictions accordingly. The number of generated feature interactions can be specified to be much smaller than the number of all the possible interactions and hence, our model admits a much lower running time. To achieve an effective algorithm, we exploit three properties of beneficial feature interactions, and propose deep-infomax-based methods to guide the interaction generation. Our experimental results show that HIRS outperforms state-of-the-art algorithms by up to 5% in terms of recommendation accuracy.

Feature interactions are essential for achieving high accuracy in recommender systems. Many studies take into account the interaction between every pair of features. However, this is suboptimal because some feature interactions may not be that relevant to the recommendation result and taking them into account may introduce noise and decrease recommendation accuracy. To make the best out of feature interactions, we propose a graph neural network approach to effectively model them, together with a novel technique to automatically detect those feature interactions that are beneficial in terms of recommendation accuracy. The automatic feature interaction detection is achieved via edge prediction with an L0 activation regularization. Our proposed model is proved to be effective through the information bottleneck principle and statistical interaction theory. Experimental results show that our model (i) outperforms existing baselines in terms of accuracy, and (ii) automatically identifies beneficial feature interactions.

User and item attributes are essential side-information; their interactions (i.e., their co-occurrence in the sample data) can significantly enhance prediction accuracy in various recommender systems. We identify two different types of attribute interactions, inner interactions and cross interactions: inner interactions are those between only user attributes or those between only item attributes; cross interactions are those between user attributes and item attributes. Existing models do not distinguish these two types of attribute interactions, which may not be the most effective way to exploit the information carried by the interactions. To address this drawback, we propose a neural Graph Matching based Collaborative Filtering model (GMCF), which effectively captures the two types of attribute interactions through modeling and aggregating attribute interactions in a graph matching structure for recommendation. In our model, the two essential recommendation procedures, characteristic learning and preference matching, are explicitly conducted through graph learning (based on inner interactions) and node matching (based on cross interactions), respectively. Experimental results show that our model outperforms state-of-the-art models. Further studies verify the effectiveness of GMCF in improving the accuracy of recommendation.

Collaborative filtering is one of the most popular techniques in designing recommendation systems, and its most representative model, matrix factorization, has been wildly used by researchers and the industry. However, this model suffers from the lack of interpretability and the item cold-start problem, which limit its reliability and practicability. In this paper, we propose an interpretable recommendation model called Multi-Matrix Factorization (MMF), which addresses these two limitations and achieves the state-of-the-art prediction accuracy by exploiting common attributes that are present in different items. In the model, predicted item ratings are regarded as weighted aggregations of attribute ratings generated by the inner product of the user latent vectors and the attribute latent vectors. MMF provides more fine grained analyses than matrix factorization in the following ways: attribute ratings with weights allow the understanding of how much each attribute contributes to the recommendation and hence provide interpretability; the common attributes can act as a link between existing and new items, which solves the item cold-start problem when no rating exists on an item. We evaluate the interpretability of MMF comprehensively, and conduct extensive experiments on real datasets to show that MMF outperforms state-of-the-art baselines in terms of accuracy.

Generative Adversarial Networks (GAN), which are capable of generating realistic synthetic real-valued data, have achieved great progress in machine learning field. However, generator in GAN framework requires being differentiable, which means that the generator cannot produce discrete data, and it poses great challenge for GAN applied in Natural Language Processing (NLP) research. To unlock the potential of GAN in NLP, we develop a novel Text-to-Text Generative Adversarial Networks (TT-GAN), through which we can get generated text based on semantic information translated from source text. We demonstrate that our model can generate not only realistic texts, but also the source text's paraphrase or its semantic summarization. As our best knowledge, it is the first framework capable of generating natural language on semantic level in real sense, and gives a new perspective to apply GAN on NLP research.