Search

Knowledge graph (KG) reasoning has been studied extensively due to its wide applications, which is addressed by two lines of research, i.e., traditional symbolic reasoning and modern neural networks-based techniques. Symbolic reasoning has been the most studied approach toward KG reasoning since the earliest days. This approach represents expert knowledge as statements in a logic-based representation language and implements reasoning as logical inference, primarily addressing high-level reasoning and cognitive processes. For instance, consider the logical rule \emph{$\text{SpeakLanguage}(x, y) \leftarrow \text{LiveIn}(x, z) \wedge \text{OfficialLanguage}(z, y)$}, along with the observed facts \emph{``LiveIn(Mina Miller, USA)''} and \emph{``OfficialLanguage(USA, English)''}, we can infer the new fact \emph{``SpeakLanguage(Mina Miller, English)''}. Symbolic reasoning approaches have demonstrated strong interpretability and generalizability due to the power of logical rules. Nevertheless, they encounter certain limitations when it comes to scaling up to vast datasets, managing ambiguity, and capturing the correlations between entities and relations. To illustrate, consider the case where both \emph{``USA''} and \emph{``United States''} denote the same country. Symbolic reasoning approaches treat them as distinct entities, thereby failing to automatically transfer the knowledge from the existing triple \emph{``LiveIn(Mina Miller, USA)''} to deduce the new fact \emph{``LiveIn(Mina Miller, United States)''}.Neural networks-based techniques have recently emerged as state-of-the-art methods in KG reasoning, owing to their successful applications. Unlike symbolic reasoning, these techniques aim to predict unseen triples by exploring the graph structure in KGs to capture the similarity of entities. Neural networks-based techniques have demonstrated good scalability and a strong ability to capture correlations between entities and relations. However, they fall short in modeling higher-order dependencies of KG relations. For instance, while neural networks-based techniques can recognize that \emph{``USA''} and \emph{``United States''} refer to the same country based on their similarity in embeddings, they cannot infer that \emph{``Mary Stilwell''} and \emph{``Mina Miller''} speak \emph{``English''} without leveraging logical rules.While symbolic reasoning and neural networks-based techniques are typically seen as distinct approaches toward achieving the ultimate goal of KG reasoning, they possess a natural potential to complement and enhance each other. Symbolic approaches offer significant interpretability and generalizability, leveraging logical rules to provide insights into inferred results and enabling easy generalization to unobserved objects. However, they struggle to handle large datasets and fail to capture correlations between entities and relations. Conversely, neural networks-based techniques address the limitations of symbolic approaches by exploring intrinsic similarities among triples, but they lack interpretability and generalizability. By integrating both techniques into a unified framework, neural-symbolic reasoning provides a more efficient, generalizable, and interpretable way to perform KG reasoning. In this thesis, we provide a comprehensive overview of techniques in integrating symbolic logical rules and neural networks for KG reasoning, with a particular focus on two specific tasks, KG completion and logical rule learning and propose our own solutions to address the limitations of existing works.

This thesis addresses the intersection of neural and symbolic artificial intelligence systems. Recent deep learning methods could memorize vast amount of world knowledge, but still have their limitation to conduct symbolic reasoning over them; while symbolic AI is good at solving reasoning tasks, but is inefficient for adapting to new knowledge. Prior efforts that bridge the two worlds mainly focus on building parsing-based systems, which require lots of annotated intermediate labels and hard to scale.My ultimate research goal is to enable neural model to interact with symbolic reasoning module in a differentiable manner, and train such Neural-Symbolic model end-to-end without intermediate labels. To bring this vision about, I have conducted works on: 1. Designing Novel Reasoning Module: design differentiable neural modules that can conduct symbolic reasoning, including knowledge graph reasoning and complex Logical inference.2. Learning via Self-Supervision: train the neural model via self-supervision from structural and symbolic knowledge base without additional annotation.3. Generalizing across Domains: the modular design of neural-symbolic system by its nature help to generalize better for Out-of-Distribution, Out-of-Vocabulary, cross-lingual and cross-type.Putting these pieces together, I am pursuing the ultimate vision to build end-to-end Neural-Symbolic system that has the capacity of reasoning, advancing to true human intelligence.

Large-scale Knowledge Graphs (KGs), such as Wikipedia and many enterprises or other domain-specific KGs, contain large numbers of real-world facts and are ubiquitous and foundational to many downstream knowledge-driven AI applications. Many existing techniques have applied state-of-the-art machine learning (ML) techniques in knowledge graph modeling to improve the performance in these applications with the KG backend, but the semantic structures especially the hierarchical ontological information inside the KGs are sparsely investigated and therefore relatively less leveraged into KG learning.In this dissertation, we demonstrate a series of research results that systematically explores how such hierarchical ontological components in knowledge graphs are incorporated into KG representation learning. We present multiple practical machine learning methods, such as hierarchical graph modeling, graph neural networks, self-supervised learning, and language models, that can effectively and efficiently capture ontological information, given different knowledge graph formulations. As a result, our proposed approaches address various real-world challenges in multiple domains, from knowledge graph itself, to diverse disciplines including natural language processing, recommender system, bioinformatics, and societal studies, and expand ML frontiers to knowledge graphs.

As an important part of Artificial Intelligence (AI), Question Answering (QA) aims at generating answers to questions in natural language. With the advancement of deep learning technology, we have witnessed substantial progress in open-domain question answering. However, QA systems are still struggling to answer questions that involve geographic entities or concepts and that require spatial operations. In order to tackle these challenges, this dissertation specifically focuses on the problem of Geographic Question Answering (GeoQA) and develops a series of spatially-explicit machine learning models to handle different GeoQA tasks. First, in Chapter 1, we discuss the challenges of answering geographic questions and the uniqueness of GeoQA. A classification of geographic questions has been presented to facilitate the development of GeoQA. Next, in Chapter 2 a spatially-explicit query relaxation model is presented to demonstrate the usefulness of geographic information and spatial thinking in the geographic question answering and query relaxation process. To develop a more generalizable approach for GeoQA and other geospatial tasks, in Chapter 3, we present a general-purpose multi-scale representation learning model for geographic locations which can be utilized in multiple downstream tasks. It has been later on utilized to build a location-aware knowledge graph embedding model for a knowledge graph-based GeoQA model in Chapter 4. Only relying on points as the spatial representations for geographic entities is not sufficient to answer many geographic questions that involve spatial relations such as topological relations and cardinal direction relations. So in Chapter 5, we present a polygon encoder that can be used to answer multiple types of spatial relation questions. In the end, we draw a conclusion by listing several challenges of GeoQA which have not been solved in this dissertation and point out some future research directions. We hope this dissertation can reveal the importance of GeoQA and demonstrate the usefulness of spatially-explicit machine learning models on geospatial problems. We also hope GeoQA will become a unique research domain and serve as an important part of Geographic Artificial Intelligent (GeoAI) research.

Over the past several years, the Web of Linked Data has continued to grow in size, both in terms of the breadth of domains covered as well as the depth and precision of knowledge. As a consequence to this growth, the community has been led to confront challenges that arise from incorporating large-scale geographic information into knowledge graphs. These challenges include data quality, data storage, data transmission, and the scaling of geospatial query processing. A crucial concern in real-time computing is about striking a balance between the time complexity of algorithms and memory consumption or data storage (i.e., space). Given a computational problem and the domain of its inputs, there are several decisions that researchers, engineers, and practitioners must make based on the constraints of available computational resources, as well as the desired program's `reaction' time for the sake of human-computer interaction. Understanding how to strike such a balance requires a thorough understanding of the data structures and algorithms used to solve a problem. Geospatial data and geospatial queries in particular require innovators to possess deep background knowledge in order to research and develop viable solutions. As a geographic information scientist working with Linked Data, I attempt to improve the quality, accessibility, reliability, and query performance of geographic data in knowledge graphs. In this dissertation, I study three specific trade-offs: (i) whether certain geographic properties and relations should be computed on-demand or materialized beforehand; (ii) whether carefully precomputing topological relations is more useful than providing users with geometries to compute topological relations on-demand; and finally, (iii) whether the challenges of hosting public geographic knowledge graph services on the Web can be mitigated, and at what cost, by a peer-to-peer architecture in which the clients possess more intelligence.

Subgraph queries are routinely used to search for entities in richly attributed graphs e.g., social networks and knowledge graphs. With little knowledge of underlying data, users often need to rewrite queries multiple times to reach desirable answers. Why-questions are studied to clarify missing or unexpected query results. This paper makes a first step to answer Why-questions for entity search in attributed graphs. We consider three common types of Why-questions: Why-not, Why, and Why-rank, which suggest query manipulations that are responsible for user-specified missing, unexpected, and undesirably ranked entities, respectively. (1) We approach a general query rewriting paradigm that suggests to identify desired entities that are specified by Why-questions. We introduce measures that characterize good query rewrites by incorporating both query editing cost and answer closeness. (2) While computing optimal query rewrites is intractable, we develop feasible algorithms, from approximation to fast heuristics, and provide query rewrites with (near) optimality guarantees whenever possible, for Why, Why-not and Why-rank questions. Using real-world graphs, we experimentally verify that our algorithms are effective and feasible for large graphs. Our case study also verifies their application in e.g., knowledge exploration.

Knowledge graphs (KGs) have become popular structures for unifying real-world entities by modelling the relationships between them and their attributes. Entity alignment -- the task of identifying corresponding entities across different KGs -- has attracted a great deal of attention in both academia and industry. However, existing alignment techniques often require large amounts of labelled data, are unable to encode multi-modal data simultaneously, and enforce only few consistency constraints. In this paper, we propose an end-to-end, unsupervised entity alignment framework for cross-lingual KGs that fuses different types of information in order to fully exploit the richness of KG data. The model captures the relation-based correlation between entities by using a multi-order graph convolutional neural (GCN) model that is designed to satisfy the consistency constraints, while incorporating the attribute-based correlation via a translation machine. We adopt a late-fusion mechanism to combine all the information together, which allows these approaches to complement each other and thus enhances the final alignment result, and makes the model more robust to consistency violations. Empirical results show that our model is more accurate and orders of magnitude faster than existing baselines. We also demonstrate its sensitivity to hyper-parameters, effort saving in terms of labelling, and the robustness against adversarial conditions.

How can data on a pre-modern book tradition be integrated into meaningful associative relations? Creating a fact-provenance driven, agile knowledge graph, we integrate data on Arabic manuscripts within their socio-historical context and present challenges and problem solutions applicable for neighbouring disciplines.

The idea of ‘citizen sensing’ and ‘human as sensors’ is crucial for social Internet of Things, an integral part of cyber-physical-social systems (CPSS). Social media data, which can be easily collected from the social world, has become a valuable resource for research in many different disciplines, e.g. crisis/disaster assessment, social event detection, or the recent COVID-19 analysis. Useful information, or knowledge derived from social data, could better serve the public if it could be processed and analyzed in more efficient and reliable ways. Advances in deep neural networks have significantly improved the performance of many social media analysis tasks. However, deep learning models typically require a large amount of labeled data for model training, while most CPSS data is not labeled, making it impractical to build effective learning models using traditional approaches. In addition, the current state-of-the-art, pre-trained Natural Language Processing (NLP) models do not make use of existing knowledge graphs, thus often leading to unsatisfactory performance in real-world applications. To address the issues, we propose a new zero-shot learning method which makes effective use of existing knowledge graphs for the classification of very large amounts of social text data. Experiments were performed on a large, real-world tweet dataset related to COVID-19, the evaluation results show that the proposed method significantly outperforms six baseline models implemented with state-of-the-art deep learning models for NLP.

Knowledge graphs have received intensive research interests. When the labels of most nodes or datapoints are missing, anchor graph and hierarchical anchor graph models can be employed. With an anchor graph or hierarchical anchor graph, we only need to optimize the labels of the coarsest anchors, and the labels of datapoints can be inferred from these anchors in a coarse-to-fine manner. The complexity of optimization is therefore reduced to a cubic cost with respect to the number of the coarsest anchors. However, to obtain a high accuracy when a data distribution is complex, the scale of this anchor set still needs to be large, which thus inevitably incurs an expensive computational burden. As such, a challenge in scaling up these models is how to efficiently estimate the labels of these anchors while keeping classification performance. To address this problem, we propose a novel approach that adds an anchor label predictor in the conventional anchor graph and hierarchical anchor graph models. In the proposed approach, the labels of the coarsest anchors are not directly optimized, and instead, we learn a label predictor which estimates the labels of these anchors with their spectral representations. The predictor is optimized with a regularization on all datapoints based on a hierarchical anchor graph, and we show that its solution only involves the inversion of a small-size matrix. Built upon the anchor hierarchy, we design a sparse intra-layer adjacency matrix over these anchors, which can simultaneously accelerate spectral embedding and enhance effectiveness. Our approach is named Faster Learning6 on Anchor Graph (FLAG) as it improves conventional anchor-graph-based methods in terms of efficiency. Experiments on a variety of publicly available datasets with sizes varying fromthousands to millions of Q1 samples demonstrate the effectiveness of our approach.

In this big data era, knowledge becomes increasingly linked, along with the rapid growth in data volume. Connected knowledge is naturally represented and stored as knowledge graphs, which are of great importance for many frontier research areas. Effectively finding relations between entities in large knowledge graphs plays a key role in many knowledge graph applications, as the most valuable part of a knowledge graph is its rich connectedness, which captures rich information about the real-world objects. However, due to the intrinsic complexity of real-world knowledge, finding semantically close relations by navigation in a large knowledge graph is challenging. Canonical graph exploration methods inevitably result in combinatorial explosion especially when the paths connecting two entities are long: the search space is $O(d^l)$ , where $d$ is the average graph node degree and l is the path length. In this paper, we will systematically study the semantic navigation problem for large knowledge graphs. Inspired by AlphaGo, which was overwhelmingly successful in game Go, we designed an efficient semantic navigation method based on a well-tailored Monte Carlo Tree Search algorithm with the unique characteristics of knowledge graphs considered. Extensive experiments show that our method is not only effective but also very efficient.

Deep Learning has been used extensively in many applications by researchers. With the increased attraction to Deep Learning, more and more unique models are created each year. However, sometimes some of the model details are not included in the publications. This makes using new Deep Learning models for research a time-consuming task for researchers. In order to tackle with this problem, we propose a prediction mechanism for the missing information in the model. By creating a dataset where the Deep Learning models are represented as knowledge graphs, we made it possible to use knowledge graph embedding algorithms which are specifically designed for eliminating missing information in a given data. We inspected 6 different algorithms and compared their performances on a small-scale experiment. After the comparison, we picked the most promising algorithm and used it for link prediction in Deep Learning models.

Knowledge graph is becoming an indispensable resource for numerous AI applications. However, the knowledge graph often suffers from its incompleteness. Building a complete, high-quality knowledge graph is time-consuming and requires significant human annotation efforts. In this paper, we study the Knowledge Graph Population task, which aims at extending the scale of structured knowledge, with a special focus on reducing data preparation and annotation efforts. Previous works mainly based on discriminative methods build classifiers and verify candidate triplets that are extracted from texts, which heavily rely on the quality of data collection and co-occurrance of entities in the text. We introduce a generative perspective to approach this task. A generative model KGGEN is proposed, which samples from the learned data distribution for each relation and can generate triplets regardless of entity pair co-occurrence in the corpus. To further improve the generation quality while alleviate human annotation efforts, adversarial learning is adopted to not only encourage generating high quality triplets, but also give model the ability to automatically assess the generation quality. Quantitative and qualitative experimental results conducted on two real-world generic knowledge graphs show that KGGEN generates novel and meaningful triplets and less human annotation comparing with the state-of-the-art approaches.

Knowledge graph link prediction uses known fact links to infer the missing link information in the knowledge graph, which is of great significance to the completion of the knowledge graph. Generating low-dimensional embeddings of entities and relations which are used to make inferences is a popular way for such link prediction problems. This paper proposes a knowledge graph link prediction method called Complex-InversE in the complex space, which maps entities and relations into the complex space. The composition of complex embeddings can handle a large variety of binary relations, among them symmetric and antisymmetric relations. The Complex-InversE effectively captures the antisymmetric relations and introduces Dropout and Early-Stopping technologies into deal with the problem of small numbers of relationships and entities, thus effectively alleviates the model’s overfitting. The results of comparison experiment on the public knowledge graph datasets show that the Complex-InversE achieves good results on multiple benchmark evaluation indicators and outperforms previous methods. Complex-InversE’s code is available on GitHub at https://github.com/ZeyuMiao97/Complex-InversE.

At present, there is a problem of insufficient utilization of regional features in the application of the existing knowledge graph based on GGNN in the defect classification of the bolt and nut pair of transmission lines. Therefore, a decision-making method of combining the bolt and nut pair with the original regional feature and the bolt and nut pair knowledge graph feature is proposed. For this reason, a method is proposed to combine the decision-making of the bolt and nut pair on the original regional features and the bolt and nut pair on the characteristics of the knowledge graph. First, the characteristics of the bolt and nut pair knowledge graph are combined with the adaptive normalization of the bolt and nut pair to the features of the original area. Then, the classification score vector based on fusion features and the classification score vector based on the bolt and nut pair to the original area feature are derived from the classifier respectively; Finally, the classification score vector of the fusion feature and the bolt and nut pair are fused to the classification score vector of the original region feature to obtain the final classification result. On this basis, this article uses bolt and nut pair to conduct multiple sets of defect classification experiments on the data set of the knowledge graph experiment. The experimental results show that the method of decision fusion of the bolt and nut pair to the original regional feature fusion the bolt and nut pair to the knowledge graph feature is better than the bolt and nut pair to the knowledge graph average precision, precision, and recall rate. It is effective Prove the improvement of the algorithm.

Modeling user’s fine-grained preferences and dynamic preference evolution from their chronological behaviors are challenging and crucial for sequential recommendation. In this paper, we develop a Hierarchical Self-Attention Incorporating Knowledge Graph for Sequential Recommendation (HSRec). HSRec models not only the user’s intrinsic preferences but also the user’s external potential interests to capture the user’s fine-grained preferences. Specifically, the intrinsic interest module and potential interest module are designed to capture these two preferences respectively. In the intrinsic interest module, user’s sequential patterns are characterized from their behaviors via the self-attention mechanism. As for the potential interest module, high-order paths can be generated with the help of the knowledge graph. Therefore, a hierarchical self-attention mechanism is designed to aggregate the semantic information of user interaction from these paths. Specifically, an entity-level self-attention mechanism is applied to capture the sequential patterns contained in the high-order paths while an interaction-level self-attention mechanism is designed to further capture the semantic information from user interactions. Moreover, according to the high-order semantic relevance, HSRec can explore the user’s dynamic preferences at each time, thus describing the user’s dynamic preference evolution. Finally, experiments conducted on three real world datasets demonstrate the state-of-the-art performance of the HSRec.

Question Answering (QA) over Knowledge Graphs (KG) aims to develop a system that is capable of answering users’ questions using the information coming from one or multiple Knowledge Graphs, like DBpedia, Wikidata, and so on. Question Answering systems need to translate the user’s question, written using natural language, into a query formulated through a specific data query language that is compliant with the underlying KG. This translation process is already non-trivial when trying to answer simple questions that involve a single triple pattern. It becomes even more troublesome when trying to cope with questions that require modifiers in the final query, i.e., aggregate functions, query forms, and so on. The attention over this last aspect is growing but has never been thoroughly addressed by the existing literature. Starting from the latest advances in this field, we want to further step in this direction. This work aims to provide a publicly available dataset designed for evaluating the performance of a QA system in translating articulated questions into a specific data query language. This dataset has also been used to evaluate three QA systems available at the state of the art.

The conventional drug discovery pipeline has proven to be unsustainable for rare diseases. Herein, we discuss recent advances in biomedical knowledge mining applied to discovering therapeutics for rare diseases. We summarize current chemogenomics data of relevance to rare diseases and provide a perspective on the effectiveness of machine learning (ML) and biomedical knowledge graph mining in rare disease drug discovery. We illustrate the power of these methodologies using a chordoma case study. We expect that a broader application of knowledge graph mining and artificial intelligence (AI) approaches will expedite the discovery of viable drug candidates against both rare and common diseases.

Urban areas have many problems, including homelessness, graffiti, and littering. These problems are influenced by various factors and are linked to each other; thus, an understanding of the problem structure is required in order to detect and solve the root problems that generate vicious cycles. Moreover, before implementing action plans to solve these problems, local governments need to estimate cost-effectiveness when the plans are carried out. Therefore, this paper proposed constructing an urban problem knowledge graph that would include urban problems' causality and the related cost information in budget sheets. In addition, this paper proposed a method for detecting vicious cycles of urban problems using SPARQL queries with inference rules from the knowledge graph. Finally, several root problems that led to vicious cycles were detected. Urban-problem experts evaluated the extracted causal relations.

In view of the lack of data association in spatiotemporal information analysis and the lack of spatiotemporal situation analysis in knowledge graphs, this paper combines the semantic web of the geographic knowledge graph with the visual analysis model of spatial information and puts forward the comprehensive utilization of the related technologies of the geographic knowledge graph and big data visual analysis. Then, it realizes the situational analysis of COVID-19 (Coronavirus Disease 2019) and the exploration of patient relationships through interactive collaborative analysis. The main contributions of the paper are as follows. (1) Based on the characteristics of the geographic knowledge graph, a patient entity model and an entity relationship type and knowledge representation method are proposed, and a knowledge graph of the spatiotemporal information of COVID-19 is constructed. (2) To analyse the COVID-19 patients’ situations and explore their relationships, an analytical framework is designed. The framework, combining the semantic web of the geographic knowledge graph and the visual analysis model of geographic information, allows one to analyse the semantic web by using the node attribute similarity calculation, key stage mining, community prediction and other methods. (3) An efficient epidemic prevention and anti-epidemic method, which is of referential significance, is proposed. It is based on experiments and the collaborative analysis of the semantic web and spatial information, allowing for real-time situational understanding, the discovery of patients’ relationships, the analysis of the spatiotemporal distribution of patients, super spreader mining, key node analysis, and the prevention and control of high-risk groups.

In the AI diagnosis of breast cancer, instead of ultrasound images from non-standard acquisition process, the Breast Image Reporting and Data System (BI-RADS) reports are widely accepted as the input data since it can give standardized descriptions for the breast ultrasound samples. The BI-RADS reports are usually stored as the format of Knowledge Graph (KG) due to the flexibility, and the KG embedding is a common procedure for the AI analysis on BI-RADS data. However, since most existing embedding methods are based on the local connections in KG, in the situation of limited labeled samples, there is a clear need for embedding based diagnosis method which is capable of representing the global interactions among all entities/relations and associating the labeled/unlabeled samples. To diagnose the breast ultrasound samples with limited labels, in this paper we propose an efficient framework Knowledge Tensor Embedding with Association Enhancement Diagnosis (KTEAED), which adopts tensor decomposition into the embedding to achieve the global representation of KG entities/relations, and introduces the association enhancement strategy to prompt the similarities between embeddings of labeled/unlabeled samples. The embedding vectors are then utilized to diagnose the clinical outcomes of samples by predicting their links to outcomes entities. Through extensive experiments on BI-RADS data with different fractions of labels and ablation studies, our KTEAED displays promising performance in the situations of various fractions of labels. In summary, our framework demonstrates a clear advantage of tackling limited labeled samples of BI-RADS reports in the breast ultrasound diagnosis.

Chemical exposure accidents pose a risk of serious injury and property damage if the diagnosis or response is not properly performed after the initial discovery. Due to lack of research on the dynamically changing environment and detection of chemical substances considering symptoms, real-time knowledge services are required, such as rapid diagnosis of chemicals exposed at the accident site and the following early response. In this study, we propose an AI-based analysis system, Symptom-based Expert for Advanced Response to Chemical Hazards (SEARCH), for chemical substance diagnosis from exposure symptoms actively collected for real-time response and mitigation to hazardous material accidents. Knowledge is collected from chemical database such as WISER, PubChem etc., and integrated for the analytics of chemical exposure accidents and contact symptoms. We design and construct ontology and knowledge graph (KG) for 1001 major chemical substances. The built KG is verified using KG embedding models and the performance of each model is compared. The proposed system identifies the substance candidates through KG query and reasoning considering the exposure conditions. Using the symptom KG, the system SEARCH can provide the means to analyze real-time data from the field and transform it into insights and actions related to emergency response.

Accurately recommending the next point of interest (POI) has become a fundamental problem with the rapid growth of location-based social networks. However, sparse, imbalanced check-in data and diverse user check-in patterns pose severe challenges for POI recommendation tasks. Knowledge-aware models are known to be primary in leveraging these problems. However, as most knowledge graphs are constructed statically, sequential information is yet integrated. In this work, we propose a meta-learned sequential-knowledge-aware recommender (Meta-SKR), which utilizes sequential, spatio-temporal, and social knowledge to recommend the next POI for a location-based social network user. The framework mainly contains four modules. First, in the graph construction module, a novel type of knowledge graph—the sequential knowledge graph, which is sensitive to the check-in order of POIs—is built to model users’ check-in patterns. To deal with the problem of data sparsity, a meta-learning module based on latent embedding optimization is then introduced to generate user-conditioned parameters of the subsequent sequential-knowledge-aware embedding module, where representation vectors of entities (nodes) and relations (edges) are learned. In this embedding module, gated recurrent units are adapted to distill intra- and inter-sequential knowledge graph information. We also design a novel knowledge-aware attention mechanism to capture information surrounding a given node. Finally, POI recommendation is provided by inferring potential links of knowledge graphs in the prediction module. Evaluations on three real-world check-in datasets show that Meta-SKR can achieve high recommendation accuracy even with sparse data.

With the global outbreak of coronavirus disease (COVID-19) all over the world, artificial intelligence (AI) technology is widely used in COVID-19 and has become a hot topic. In recent 2 years, the application of AI technology in COVID-19 has developed rapidly, and more than 100 relevant papers are published every month. In this paper, we combined with the bibliometric and visual knowledge map analysis, used the WOS database as the sample data source, and applied VOSviewer and CiteSpace analysis tools to carry out multi-dimensional statistical analysis and visual analysis about 1903 pieces of literature of recent 2 years (by the end of July this year). The data is analyzed by several terms with the main annual article and citation count, major publication sources, institutions and countries, their contribution and collaboration, etc. Since last year, the research on the COVID-19 has sharply increased; especially the corresponding research fields combined with the AI technology are expanding, such as medicine, management, economics, and informatics. The China and USA are the most prolific countries in AI applied in COVID-19, which have made a significant contribution to AI applied in COVID-19, as the high-level international collaboration of countries and institutions is increasing and more impactful. Moreover, we widely studied the issues: detection, surveillance, risk prediction, therapeutic research, virus modeling, and analysis of COVID-19. Finally, we put forward perspective challenges and limits to the application of AI in the COVID-19 for researchers and practitioners to facilitate future research on AI applied in COVID-19.

Consumption of media, and movies in particular, is increasing and is influenced by a number of factors. One important and overlooked factor that affects the media consumption choices is the emotional state of the user and the decision making based on it. To include this factor in movie recommendation processes, we propose a knowledge graph representing human emotions in the domain of movies. The knowledge graph has been built by extracting emotions out of pre-existing movie reviews using machine learning techniques. To show how the knowledge graph can be used, a chatbot prototype has been developed. The chatbot’s reasoning mechanism derives movie recommendations for the user by combining the user’s emotions, which have been extracted from chat messages, with the knowledge graph. The developed approach for movie recommendations based on sentiment represented as a knowledge graph has been proven to be technically feasible, however, it requires more information about the emotions associated with the movies than currently available online.

Few‐shot knowledge graph (KG) reasoning is the main focus in the field of knowledge graph reasoning. In order to expand the application fields of the knowledge graph, a large number of studies are based on a large number of training samples. However, we have learnt that there are actually many missing relationships or entities in the knowledge graph, and in most cases, there are not many training instances when implementing new relationships. To tackle it, in this study, the authors aim to predict a new entity given few reference instances, even only one training instance. A few‐shot learning framework based on a hybrid attention mechanism is proposed. The framework employs traditional embedding models to extract knowledge, and uses an attenuated attention network and a self‐attention mechanism to obtain the hidden attributes of entities. Thus, it can learn a matching metric by considering both the learnt embeddings and one‐hop graph structures. The experimental results present that the model has achieved significant performance improvements on the NELL‐One and Wiki‐One datasets.

An organized knowledge base can facilitate the exploration of existing knowledge and the detection of emerging topics in a domain. Knowledge about and around Geographic Information Science and its associated system technologies (GIS) is complex, extensive and emerging rapidly. Taking the challenge, we built a GIS knowledge graph (GIS-KG) by (1) merging existing GIS bodies of knowledge to create a hierarchical ontology and then (2) applying deep-learning methods to map GIS publications to the ontology. We conducted several experiments on information retrieval to evaluate the novelty and effectiveness of the GIS-KG. Results showed the robust support of GIS-KG for knowledge search of existing GIS topics and potential to explore emerging research themes.

Cross-lingual entity alignment has attracted considerable attention in recent years. Past studies using conventional approaches to match entities share the common problem of missing important structural information beyond entities in the modeling process. This allows graph neural network models to step in. Most existing graph neural network approaches model individual knowledge graphs (KGs) separately with a small amount of pre-aligned entities served as anchors to connect different KG embedding spaces. However, this characteristic can cause several major problems, including performance restraint due to the insufficiency of available seed alignments and ignorance of pre-aligned links that are useful in contextual information in-between nodes. In this article, we propose DuGa-DIT, a dual gated graph attention network with dynamic iterative training, to address these problems in a unified model. The DuGa-DIT model captures neighborhood and cross-KG alignment features by using intra-KG attention and cross-KG attention layers. With the dynamic iterative process, we can dynamically update the cross-KG attention score matrices, which enables our model to capture more cross-KG information. We conduct extensive experiments on two benchmark datasets and a case study in cross-lingual personalized search. Our experimental results demonstrate that DuGa-DIT outperforms state-of-the-art methods.

The rapid growth of research publications has placed great demands on digital libraries (DL) for advanced information management technologies. To cater to these demands, techniques relying on knowledge-graph structures are being advocated. In such graph-based pipelines, inferring semantic relations between related scientific concepts is a crucial step. Recently, BERT-based pre-trained models have been popularly explored for automatic relation classification. Despite significant progress, most of them were evaluated in different scenarios, which limits their comparability. Furthermore, existing methods are primarily evaluated on clean texts, which ignores the digitization context of early scholarly publications in terms of machine scanning and optical character recognition (OCR). In such cases, the texts may contain OCR noise, in turn creating uncertainty about existing classifiers' performances. To address these limitations, we started by creating OCR-noisy texts based on three clean corpora. Given these parallel corpora, we conducted a thorough empirical evaluation of eight Bert-based classification models by focusing on three factors: (1) Bert variants; (2) classification strategies; and, (3) OCR noise impacts. Experiments on clean data show that the domain-specific pre-trained Bert is the best variant to identify scientific relations. The strategy of predicting a single relation each time outperforms the one simultaneously identifying multiple relations in general. The optimal classifier's performance can decline by around 10% to 20% in F-score on the noisy corpora. Insights discussed in this study can help DL stakeholders select techniques for building optimal knowledge-graph-based systems.

Matrix Factorization (MF) is a widely used collaborative filtering technique for effectively modeling a user-item interaction in recommender system. Despite the successful application of MF and its variants, the method proves to be effective only in situations where there is an abundance of user-item interactions. However, user-item interaction data are usually sparse, limiting the effectiveness of the method. In addressing this problem, recent methods have proposed to use knowledge graphs (KGs) as additional information to complement the sparse user-item interaction data. This has proved challenging given the complexity of the KG structure. In this paper, we propose a collaborative filtering method that takes advantage of knowledge graphs. More specifically, the embedding of a user and item are both grounded on the item’s attributes in the knowledge graph, and are aggregated with generic user and item representations modeled by MF for implicit recommendation. Our model has demonstrated to outperform the recent state-of-the-art method KGCN [18] in very sparse settings, showing an effective integration of KGs in recommender systems.

Since an increasing number of applications have urgent requirements for the management of fuzzy spatiotemporal information, it is challenging to model fuzzy spatiotemporal data . On the other hand, because Resource Description Framework (RDF) benefits greatly from web data, linked data , and knowledge graph, it has been widely accepted and has rapidly become a new choice to model and represent data in many application domains. The advent of RDF seems to provide an opportunity for dealing with fuzzy spatiotemporal data. However, RDF data is subject to continuously changing operations so that they may violate the integrity constraints and cause inconsistencies. In order to solve these problems, we propose an extended RDF model that can represent fuzzy spatiotemporal data, and then present the structure of fuzzy spatiotemporal data in RDF graph. After studying the constraint conditions of fuzzy spatiotemporal data in RDF graph, we propose the corresponding fixing rules and algorithms to detect and repair the fuzzy spatiotemporal RDF graph. Finally, the experimental results demonstrate the effectiveness of our approach which can significantly solve the inconsistency problem of fuzzy spatiotemporal RDF graphs.

Knowledge graphs (KG) often encounter knowledge incompleteness. The path reasoning that predicts the unknown path relation between pairwise entities based on existing facts is one of the most promising approaches to the knowledge graph completion. However, most conventional path reasoning methods exclusively consider the entity description included in fact triples, ignoring both the type information of entities and the interaction between different semantic representations. In this study, we propose a novel method, Type-aware Attentive Path Reasoning (TAPR), to complete the knowledge graph by simultaneously considering KG structural information, textual information, and type information. More specifically, we first leverage types to enrich the representational learning of entities and relationships. Next, we describe a type-level attention to select the most relevant type of given entity in a specific triple without any predefined rules or patterns to reduce the impact of noisy types. After learning the distributed representation of all paths, path-level attention assigns different weights to paths, from which relations among entity pairs are calculated. We conduct a series of experiments on a real-world dataset to demonstrate the effectiveness of TAPR. Experimental results show that our method significantly outperforms all baselines on link prediction and entity prediction tasks.

Temporal knowledge graph (TKG) representation learning embeds relations and entities into a continuous low-dimensional vector space by incorporating temporal information. Latest studies mainly aim at learning entity representations by modeling entity interactions from the neighbor structure of the graph. However, the interactions of relations from the neighbor structure of the graph are neglected, which are also of significance for learning informative representations. In addition, there still lacks an effective historical relation encoder to model the multi-range temporal dependencies. In this article, we propose a d ual gr a ph c onvolution network based TKG representation learning method using h istorical rel a tions (DACHA). Specifically, we first construct the primal graph according to historical relations, as well as the edge graph by regarding historical relations as nodes. Then, we employ the dual graph convolution network to capture the interactions of both entities and historical relations from the neighbor structure of the graph. In addition, the temporal self-attentive historical relation encoder is proposed to explicitly model both local and global temporal dependencies. Extensive experiments on two event based TKG datasets demonstrate that DACHA achieves the state-of-the-art results.

The intelligent inquiry system for metro electro-mechanical equipment faults based on the knowledge graph can effectively consolidate various semi-structured failure messages, and can provide users with quick, accurate and high-quality intelligent inquiry services such as equipment fault causes-researching and solutions-delivering, which could be really relevant to this research field and application areas. The recorded date which related to metro electromechanical equipment failures were in this research collected, consolidated and converted, so that these failures could be stored in our databases. In this context, various functions of the intelligent inquiry system have been implemented, including: natural language question analysis, language Cypher-based question and answer design, Naive Bayesian classification based on characteristic core words, and user interaction interface realization. The experimental results show that the system can effectively solve the problems related to fault handling in metro mechanical and electrical equipment, thus improving the efficiency of equipment fault maintenance.

In recent years, studies have revealed that introducing knowledge graphs (KGs) into recommendation systems as auxiliary information can improve recommendation accuracy. However, KGs are usually based on third-party data that may be manipulated by malicious individuals. In this study, we developed a poisoning attack strategy applied on a KG-based recommendation system to analyze the influence of fake links. The aim of an attacker is to recommend specific products to improve their visibility. Most related studies have focused on adversarial attacks on graph data; KG-based recommendation systems have rarely been discussed. We propose an attack model corresponding to recommendations. In the model, the current recommended status and a specified item are analyzed to estimate the effects of different attack decisions (addition or deletion of facts), thereby generating the optimal attack combination. Finally, the KG is contaminated by the attack combination so that the trained recommendation model recommends a specific item to as many people as possible. We formulated the process into a deep reinforcement learning method. Conducting experiments on the movie and the fund data sets enabled us to systematically analyze our poisoning attack strategy. The experimental results proved that the proposed strategy can effectively improve an item’s ranking in a recommendation list.