Article
Computer Science, Information Systems
Ibrahim A. Ahmed, Fatima N. AL-Aswadi, Khaled M. G. Noaman, Wafa' Za'al Alma'aitah
Summary: With the growth of data on the Web, the need for efficient methods to extract valuable information from the data has increased. Knowledge graphs provide an efficient and easy way to represent and organize data. The construction of Arabic Knowledge Graph (AKG) faces challenges due to limited Arabic data and lack of effective language processing tools. This research reviews KG construction best practices and discusses the challenges and potential solutions in constructing AKG.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Wenqiang Liu, Hongyun Cai, Xu Cheng, Sifa Xie, Yipeng Yu, Dukehyzhang
Summary: The goal of representation learning of knowledge graph is to encode entities and relations into a low-dimensional embedding space. Existing methods have limitations in expressing high-order structural relationships between entities and utilizing attribute triples. To overcome these limitations, this paper proposes a novel method named KANE, which captures high-order structural and attribute information of knowledge graphs using graph convolutional networks. Experimental results show that KANE outperforms other methods.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Zhen Bi, Siyuan Cheng, Jing Chen, Xiaozhuan Liang, Feiyu Xiong, Ningyu Zhang
Summary: In this paper, we propose a new variant of Transformer called Relphormer for knowledge graph representations. We introduce Triple2Seq to dynamically sample contextualized sub-graph sequences as input, alleviating the heterogeneity issue. We also propose a novel structure-enhanced self-attention mechanism to encode relational information. Experimental results show that Relphormer outperforms baseline models.
Article
Computer Science, Artificial Intelligence
Weihang Zhang, Ovidiu Serban, Jiahao Sun, Yike Guo
Summary: This paper proposes a novel method for multiple knowledge graph completion by leveraging information from other knowledge graphs to alleviate the sparseness of a single knowledge graph, achieving state-of-the-art results on multilingual knowledge graph datasets.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Yinghan Shen, Xuhui Jiang, Zijian Li, Yuanzhuo Wang, Chengjin Xu, Huawei Shen, Xueqi Cheng
Summary: In this paper, a unified social knowledge graph representation learning framework (UniSKGRep) is proposed to improve the downstream tasks of user modeling by leveraging the multi-view information inherent in the social network (SN) and knowledge graph (KG). Extensive experiments demonstrate that UniSKGRep achieves general and substantial performance improvement compared to strong baselines in various downstream tasks.
Article
Computer Science, Artificial Intelligence
Qian Li, Daling Wang, Shi Feng, Kaisong Song, Yifei Zhang, Ge Yu
Summary: In this paper, a new method called Open knowledge graph Enhanced Representation Learning of KGs (OERL) is proposed to address the sparseness and incompleteness of knowledge graphs. OERL extracts textual and structural connections between knowledge graphs and open knowledge graphs to enhance the representation learning. Experimental results demonstrate that OERL outperforms state-of-the-art baselines.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Yujia Huo, Derek F. Wong, Lionel M. Ni, Lidia S. Chao, Jing Zhang, Xin Zuo
Summary: Measuring learner cognition based on their problem-solving performance is a joint discipline of cognitive psychology and machine learning. By using signed knowledge interaction network, it can better capture complete cognition-mis-cognition proximity information. The learned knowledge embedding can achieve learner performance prediction tasks and show promising prediction scores compared to several methods in network sign prediction and learner performance prediction.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Qian Li, Daling Wang, Shi Feng, Kaisong Song, Yifei Zhang, Ge Yu
Summary: This paper proposes the VGAT model and CR protocol to address the prediction of missing links in open knowledge graphs. The VGAT model automatically mines synonymous features using a variational autoencoder densified graph attention mechanism, while the CR protocol comprehensively evaluates multiple answers from the perspectives of significance and compactness.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Automation & Control Systems
Shih-Yuan Yu, Sujit Rokka Chhetri, Arquimedes Canedo, Palash Goyal, Mohammad Abdullah Al Faruque
Summary: Pykg2vec is a Python library for learning representations of entities and relations in knowledge graphs, implementing 25 state-of-the-art knowledge graph embedding algorithms and designed to accelerate research in knowledge graph representation learning. Released under the MIT License, Pykg2vec is built on PyTorch and Python's multiprocessing framework.
JOURNAL OF MACHINE LEARNING RESEARCH
(2021)
Article
Computer Science, Artificial Intelligence
Daiyi Li, Li Yan, Xiaowen Zhang, Wei Jia, Zongmin Ma
Summary: This paper verifies the importance of embedding event knowledge in KG representation learning and proposes a novel event KGE model based on event causal transfer. The model effectively maintains the semantic information of events, entities, and relations in the event KG through a six tuple-based event representation model and a constructed heterogeneous graph. Information transfer based on an attention network is used to integrate event information into KGE. Extensive experiments demonstrate the efficiency and stability of the designed event KGE model in multiple downstream tasks.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Peng He, Gang Zhou, Mengli Zhang, Jianghong Wei, Jing Chen
Summary: The approach of knowledge graph embedding enables the representation of facts in low-dimensional vector spaces, reducing complexity. However, most approaches ignore the time attribute. To address this, we propose two temporal KGE models that integrate time information and can handle common challenges in real-world temporal KGs.
APPLIED INTELLIGENCE
(2023)
Article
Green & Sustainable Science & Technology
Xuan Guo, Haizhong Qian, Fang Wu, Junnan Liu
Summary: GeoKG is a geographical knowledge graph constructed based on multisource data, with a modeling schema layer and a filling data layer for knowledge extraction, integration, and storage. Experimental results show that GeoKG can automatically extract and integrate knowledge from multisource data, achieving a high success rate and 100% exact coordinates.
Article
Computer Science, Artificial Intelligence
Haichuan Fang, Youwei Wang, Zhen Tian, Yangdong Ye
Summary: This study introduces an innovative framework called D-AEN, which propagates and updates the representations of both relations and entities by fusing neighborhood information. It enables elements like relations and entities to interact well semantically, thereby retaining more effective information of knowledge graphs.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Zongcai Huang, Peiyuan Qiu, Li Yu, Feng Lu
Summary: Geographic relation completion is crucial for improving the quality of large-scale geographic knowledge graphs (GeoKGs). This study proposes a geographic relation prediction model based on multi-layer similarity enhanced networks, which effectively enhances the implicit semantic associations between existing geographic entities. Experimental results demonstrate the high accuracy of this method in geo-relations prediction.
ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION
(2022)
Article
Computer Science, Information Systems
Xin Wang, Shengfei Lyu, Xiangyu Wang, Xingyu Wu, Huanhuan Chen
Summary: Knowledge Graph Completion (KGC) is a fundamental problem for temporal knowledge graphs (TKGs), and existing TKG embedding methods face scalability issues and lack of global information utilization. To address these issues, we propose a novel and effective TKG embedding method called Temporal Knowledge Graph Embedding via Sparse Transfer Matrix (TASTER), which combines global and local information. TASTER learns global embeddings from a static knowledge graph and derives local embeddings from global embeddings based on specific subgraphs. It also utilizes sparse transformation matrices to adapt to TKGs with varying subgraphs. Experimental results on real-world datasets demonstrate that TASTER outperforms existing models in TKG link prediction tasks, validating its effectiveness.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Information Systems
Peng Ye, Xueying Zhang, Chunju Zhang, Yulong Dang
Summary: This paper proposes a positioning method based on supervaluation semantics to address the vagueness in location description in different contexts. By analyzing human spatial cognition and the types of attention elements in location description, a positioning model is constructed. A question-answering system is designed for data collection and a case study is conducted to verify the effectiveness of the method.
ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION
(2022)
Article
Geography
Qinjun Qiu, Zhong Xie, Shu Wang, Yunqiang Zhu, Hairong Lv, Kai Sun
Summary: This article proposes a weakly supervised Chinese toponym recognition (ChineseTR) architecture, which automatically generates training datasets and utilizes a bidirectional recurrent neural network for toponym recognition. Experimental results show that ChineseTR achieves good performance in toponym recognition.
TRANSACTIONS IN GIS
(2022)
Article
Environmental Sciences
Peng Ye
Summary: Meteorological disaster monitoring is an important application of remote sensing technology in the field of meteorology. The key issues include task arrangement and organization, information extraction, and change detection. Future research directions include sensor planning, information model construction, and multi-source data fusion to promote process monitoring of meteorological disasters.
INTERNATIONAL JOURNAL OF ENVIRONMENTAL RESEARCH AND PUBLIC HEALTH
(2022)
Article
Chemistry, Multidisciplinary
Mengfei Xu, Shu Wang, Chenlong Song, Anqi Zhu, Yunqiang Zhu, Zhiqiang Zou
Summary: This paper proposes a geographic data augmentation method called S-GANs, which combines generative adversarial networks (GAN) with the Third Law of Geography, to address the issue of sparse rural attribute data and improve the accuracy of recommendation algorithms.
APPLIED SCIENCES-BASEL
(2022)
Review
Chemistry, Multidisciplinary
Peng Ye, Guowei Liu, Yi Huang
Summary: High-spatial-resolution remote sensing images are crucial for describing the details of objects and understanding the semantic information and relationships in geographic scenes. This paper summarizes the key factors and processing strategies in geographic scene understanding and discusses future challenges and prospects.
APPLIED SCIENCES-BASEL
(2022)
Article
Multidisciplinary Sciences
Wei Tang, Xueying Zhang, Deen Feng, Yipeng Wang, Peng Ye, Hanhua Qu
Summary: In this study, the components and characteristics of alpine skiing were analyzed, and a knowledge graph of alpine skiing was constructed using multi-source data. The proposed method allows for clear representation of the comprehensive relationships between meteorological conditions and alpine skiing, providing decision support for event management.
Review
Environmental Sciences
Shu Wang, Yunqiang Zhu, Lang Qian, Jia Song, Wen Yuan, Kai Sun, Weirong Li, Quanying Cheng
Summary: This paper proposes a novel and rapid approach called WEAPI for investigating ecological agricultural patterns based on web-text. The proposed method achieves high precision and coverage rates in detecting trends in Chinese ecological agriculture, making it a promising and powerful tool for agricultural research and development planning.
SCIENCE OF THE TOTAL ENVIRONMENT
(2022)
Article
Geosciences, Multidisciplinary
Shu Wang, Yunqiang Zhu, Yanmin Qi, Zhiwei Hou, Kai Sun, Weirong Li, Lei Hu, Jie Yang, Hairong Lv
Summary: Time is a fundamental reference system in geosciences for recording and interpreting temporal information. However, there are limitations in the current time conversion method due to the scope of existing time ontologies and reliance on experience. To address these issues, this paper proposes a Unified Time Framework (UTF) that designs an independent time root node and incorporates quantitative relationship definitions, unified time nodes, and interfaces to enhance accuracy and efficiency in calculating temporal information across different time references. Experimental results demonstrate that UTF greatly supports accurate and efficient temporal information queries compared to traditional time conversion methods. The UTF can be widely applied in the era of Big Data to discover new geosciences knowledge across different time references.
GEOSCIENCE FRONTIERS
(2023)
Article
Remote Sensing
Kehan Wu, Xueying Zhang, Yulong Dang, Peng Ye
Summary: This paper investigates the performance of pre-trained language model-based spatial relation extraction methods on Chinese text, compares the differences between the pipeline extraction and joint extraction approaches, and finds that the pipeline extraction method outperforms the joint extraction method for Chinese text data.
GEO-SPATIAL INFORMATION SCIENCE
(2023)
Article
Computer Science, Information Systems
Xuhui Zeng, Shu Wang, Yunqiang Zhu, Mengfei Xu, Zhiqiang Zou
Summary: The recommendation system for countryside ecological patterns is an important application of artificial intelligence in rural development. However, existing methods suffer from low accuracy due to the neglect of complex geographical features. To address this issue, this study proposes a geographical Knowledge Graph Convolutional Networks method, which achieves higher accuracy by addressing data sparsity and the "cold start" problem.
ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION
(2022)
Article
Geography
Jinqu Zhang, Lang Qian, Shu Wang, Yunqiang Zhu, Zhenji Gao, Hailong Yu, Weirong Li
Summary: In this study, a corpus augmentation method based on Levenshtein distance was proposed to enrich the geoscience corpus by constructing a geoscience dictionary and calculating the distance between words. A Chinese word segmentation model combining BERT, Bi-GRU, and CRF was implemented. Experimental results showed significant performance improvement of the proposed method, indicating great potential for natural language processing tasks like named entity recognition and relation extraction.
Article
Environmental Sciences
Xiaoliang Dai, Yunqiang Zhu, Kai Sun, Qiang Zou, Shen Zhao, Weirong Li, Lei Hu, Shu Wang
Summary: Landslide susceptibility assessment in Liangshan, China was investigated using the geographical random forest (GRF) model. Compared to random forest (RF), GRF achieved higher performance with an AUC of 0.86 by considering spatial heterogeneity among variables. GRF also provided a higher-quality landslide susceptibility map, correctly identifying 92.35% of landslide points in high-susceptibility areas. The local feature importance derived from GRF revealed spatial variation in the impact of conditioning factors, providing implications for policy development to prevent and mitigate landslides.
Article
Green & Sustainable Science & Technology
Peng Ye, Yuping Liu
Summary: With the development of the mobile Internet, new media platforms play a crucial role in assisting tourist attractions to provide tourism services. This study focuses on the tourist attractions in Yangzhou, China and proposes a classification system and indicator system to evaluate the construction level of new media platforms. The findings highlight the differences among platforms and types of functions, as well as areas for improvement.
Review
Computer Science, Information Systems
Hao Sun, Shu Wang, Yunqiang Zhu, Wen Yuan, Zhiqiang Zou
Summary: In this paper, a comprehensive question classification framework (IQA_QC) is proposed to accurately understand user query intention in Geospatial Intelligent Question Answering (GeoIQA). The framework covers the complexity and diversity of geographical questions by introducing the basic idea of the IQA mechanism. However, there are significant deficiencies in the current IQA evaluation metrics in broader dimensions. In comparison, the proposed IQA_QC framework can integrate and surpass existing classification.
ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION
(2023)
Article
Computer Science, Information Systems
Jie Zhu, Ziqi Lang, Shu Wang, Mengyao Zhu, Jiaming Na, Jiazhu Zheng
Summary: This study selected three popular sources of night-time light data to identify the urban fringe and employed three representative dual spatial clustering approaches for extracting urban fringe areas. The study found that NASA's Black Marble data provided a reliable approach for accurately extracting urban fringe areas.
ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION
(2023)