Article
Computer Science, Information Systems
Fei Luo, Salabat Khan, Yandao Huang, Kaishun Wu
Summary: Wearable devices with various sensors can measure physiological and behavioral characteristics. Activity-based person identification is a growing technology in security and access control. Smartphones, Apple Watch, and Google Glass can collect activity-related information for differentiating individuals. This article implemented eight classifiers, including MSENet, TST, TCN, CNN-LSTM, ConvLSTM, XGBoost, decision tree, and k-nearest neighbor, achieving high person identification accuracies on public datasets.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Review
Biochemical Research Methods
Soren Richard Stahlschmidt, Benjamin Ulfenborg, Jane Synnergren
Summary: Biomedical data are increasingly multimodal, and deep learning-based data fusion strategies are effective in capturing their complex relationships, especially joint representation learning which models the interactions between different levels of biological organization.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Genetics & Heredity
Yan Wang, Rui Guo, Lan Huang, Sen Yang, Xuemei Hu, Kai He
Summary: N-6-methyladenosine (m(6)A) is a prevalent RNA post-transcriptional modification with significant biological implications. The proposed predictor m(6)AGE combines sequence-derived and graph embedding features for m(6)A site prediction, outperforming other predictors across multiple datasets.
FRONTIERS IN GENETICS
(2021)
Article
Mathematics, Interdisciplinary Applications
N. Nikzad-Khasmakhi, M. A. Balafar, M. Reza Feizi-Derakhshi, Cina Motamed
Summary: The study introduces a recommendation system approach called BERTERS that incorporates expert candidate scores such as knowledge level, reputation, and influence into a single vector representation using BERT and graph embedding techniques. This approach significantly improves recommendation accuracy and performance across various tasks and classifiers, demonstrating its potential in diverse domains.
CHAOS SOLITONS & FRACTALS
(2021)
Review
Computer Science, Artificial Intelligence
Weipeng Cao, Yuhao Wu, Yixuan Sun, Haigang Zhang, Jin Ren, Dujuan Gu, Xingkai Wang
Summary: Multimodal learning allows for the utilization of various types of information to provide a comprehensive view for modeling targets. Zero-shot learning integrates prior knowledge into data-driven models for accurate class identification. Combining these two approaches, known as multimodal zero-shot learning, can harness the advantages of both and lead to models with better generalization abilities. However, comprehensive research and summaries on multimodal zero-shot learning algorithms and applications are currently lacking. This study aims to bridge this gap by offering an objective overview of the definition, typical algorithms, representative applications, and crucial issues surrounding multimodal zero-shot learning. This article not only provides researchers in this field with a comprehensive perspective but also highlights several promising research directions.
WILEY INTERDISCIPLINARY REVIEWS-DATA MINING AND KNOWLEDGE DISCOVERY
(2023)
Article
Computer Science, Artificial Intelligence
Chi Thang Duong, Thanh Tam Nguyen, Trung-Dung Hoang, Hongzhi Yin, Matthias Weidlich, Quoc Viet Hung Nguyen
Summary: We present Deep MinCut (DMC), an unsupervised approach for learning node embeddings in graph-structured data. DMC derives node representations based on their membership in communities, eliminating the need for a separate clustering step. By minimizing the mincut loss, which captures connections between communities, DMC learns both node embeddings and communities simultaneously. Our empirical evidence demonstrates that the communities learned by DMC are meaningful and that the node embeddings perform well in various node classification benchmarks.
PATTERN RECOGNITION
(2022)
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
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
Mathematical & Computational Biology
Shi Liu, Kaiyang Li, Yaoying Wang, Tianyou Zhu, Jiwei Li, Zhenyu Chen
Summary: Knowledge graph embedding aims to learn representation vectors for entities and relations. Existing approaches mainly use structural information to learn the representation, neglecting content related to entities and relations. In this paper, a multi-modal content fusion model is proposed to effectively fuse heterogeneous data for knowledge graph embedding. Experimental results show that the proposed model outperforms state-of-the-art methods significantly, indicating its superiority.
MATHEMATICAL BIOSCIENCES AND ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Piotr Bielak, Kamil Tagowski, Maciej Falkiewicz, Tomasz Kajdanowicz, Nitesh V. Chawla
Summary: This paper discusses the challenges of representation learning on dynamic graphs and proposes a framework called FILDNE for incorporating advances in static representation learning into dynamic graphs. FILDNE reduces computational costs while improving quality measure gains by applying static representation learning methods to dynamic graphs.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Xue Liu, Wei Wei, Xiangnan Feng, Xiaobo Cao, Dan Sun
Summary: The paper introduces a novel graph embedding algorithm named GraphCSC, which integrates skeleton information and component information of graphs into embeddings for classification. Experiments demonstrate that the algorithm outperforms state-of-the-art baselines in graph classification tasks.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Piotr Bielak, Tomasz Kajdanowicz, Nitesh V. Chawla
Summary: Self-supervised learning is an important area of research that aims to eliminate the need for expensive data labeling. We propose a new framework called Graph Barlow Twins for self-supervised graph representation learning, which utilizes a cross-correlation-based loss function instead of negative samples.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Theory & Methods
Ryan Rivas, Sudipta Paul, Vagelis Hristidis, Evangelos E. Papalexakis, Amit K. Roy-Chowdhury
Summary: Twitter is a frequently used subject in machine learning research and applications, with problems like sentiment analysis, image tagging, and location prediction being studied on Twitter data. Most prior work in this area focuses on a subset of the available data, such as text or text and image. However, a tweet can have additional components like location and author, which can provide useful information for machine learning tasks. In this study, the authors propose a deep neural network framework that combines text, image, and graph representations to learn joint embeddings for different tweet components. Experimental results show that this approach has comparable or superior performance compared to specialized application-specific approaches.
JOURNAL OF BIG DATA
(2022)
Article
Computer Science, Artificial Intelligence
Alexis Cvetkov-Iliev, Alexandre Allauzen, Gael Varoquaux
Summary: For many machine learning tasks, improving performance requires augmenting the data table with features derived from external sources. This study proposes replacing manually crafted features with vector representations of entities, such as cities, that capture relevant information. The research shows the importance of modeling entity relationships and capturing numerical attributes for creating effective feature vectors from relational data.
Article
Computer Science, Artificial Intelligence
Lijun Dong, Hong Yao, Dan Li, Yi Wang, Shengwen Li, Qingzhong Liang
Summary: Graph embedding technique in artificial intelligence is important for processing complex graph data efficiently. Existing GNN models often have limitations in considering global topology information, leading to difficulties in distinguishing nodes with similar local topologies. The proposed AS-GNN model aims to address this issue by capturing global topology information based on the characteristics of complex networks.
KNOWLEDGE-BASED SYSTEMS
(2021)