Open Access. Powered by Scholars. Published by Universities.®
- Discipline
-
- Databases and Information Systems (3560)
- Software Engineering (2205)
- Artificial Intelligence and Robotics (1897)
- Information Security (1108)
- Numerical Analysis and Scientific Computing (1060)
-
- Graphics and Human Computer Interfaces (947)
- Engineering (884)
- Social and Behavioral Sciences (808)
- Business (748)
- Theory and Algorithms (513)
- Computer Engineering (449)
- Programming Languages and Compilers (413)
- Operations Research, Systems Engineering and Industrial Engineering (407)
- OS and Networks (345)
- Communication (326)
- Social Media (264)
- Public Affairs, Public Policy and Public Administration (230)
- Medicine and Health Sciences (197)
- Education (194)
- Transportation (194)
- Management Information Systems (176)
- Data Storage Systems (167)
- E-Commerce (154)
- International and Area Studies (147)
- Technology and Innovation (146)
- Asian Studies (145)
- Health Information Technology (118)
- Higher Education (105)
- Keyword
-
- Machine learning (145)
- Deep learning (129)
- Artificial intelligence (124)
- Social media (82)
- Singapore (73)
-
- Reinforcement learning (72)
- Data mining (70)
- Privacy (67)
- Security (62)
- Cloud computing (60)
- Deep Learning (58)
- Empirical study (55)
- Software engineering (55)
- Optimization (54)
- Online learning (51)
- Visualization (51)
- Neural networks (50)
- Anomaly detection (49)
- Training (49)
- Twitter (49)
- Task analysis (48)
- Blockchain (47)
- Large Language Models (47)
- Natural language processing (47)
- Collaboration (46)
- Feature extraction (45)
- Algorithms (44)
- Access control (43)
- Machine Learning (43)
- Semantics (43)
- Publication Year
- Publication
-
- Research Collection School Of Computing and Information Systems (8479)
- Dissertations and Theses Collection (Open Access) (189)
- Research Collection Lee Kong Chian School Of Business (59)
- Research Collection Yong Pung How School Of Law (49)
- Research Collection School of Social Sciences (27)
-
- Asian Management Insights (26)
- Research Collection College of Integrative Studies (23)
- Perspectives@SMU (21)
- Research Collection School Of Accountancy (18)
- Dissertations and Theses Collection (15)
- FORCE 2026 (14)
- SMU Press Releases and News (12)
- MITB Thought Leadership Series (11)
- Research Collection School of Computing and Information Systems (11)
- Research Collection Library (10)
- Research@SMU: Connecting the Dots (10)
- PhD Student’s Publications Collection (8)
- LARC Research Publications (7)
- Research Collection School Of Economics (6)
- CCX Research (4)
- SMU Research Data (4)
- Student Publications (4)
- 2024 AI for Research Week (3)
- SCIS Student Publications (3)
- Centre for AI & Data Governance (2019-2025) (2)
- Research Collection Office of Research (2)
- CASTLe: Collection of Articles on Scholarship for Teaching and Learning (1)
- Centre for Computational Law (2022-2025) (1)
- Library Events (1)
- ROSA Journal Articles and Publications (1)
- Publication Type
- File Type
Articles 3991 - 4020 of 9025
Full-Text Articles in Computer Sciences
Decision Making For Improving Maritime Traffic Safety Using Constraint Programming, Saumya Bhatnagar, Akshat Kumar, Hoong Chuin Lau
Decision Making For Improving Maritime Traffic Safety Using Constraint Programming, Saumya Bhatnagar, Akshat Kumar, Hoong Chuin Lau
Research Collection School Of Computing and Information Systems
Maritime navigational safety is of utmost importance to prevent vessel collisions in heavily trafficked ports, and avoid environmental costs. In case of a likely near miss among vessels, port traffic controllers provide assistance for safely navigating the waters, often at very short lead times. A better strategy is to avoid such situations from even happening. To achieve this, we a) formalize the decision model for traffic hotspot mitigation including realistic maritime navigational features and constraints through consultations with domain experts; and b) develop a constraint programming based scheduling approach to mitigate hotspots. We model the problem as a variant of …
Data-Driven Decision-Support For Process Improvement Through Predictions Of Bed Occupancy Rates, Kar Way Tan, Qi You Ng, Francis Ngoc Hoang Long Nguyen, Sean Shao Wei Lam
Data-Driven Decision-Support For Process Improvement Through Predictions Of Bed Occupancy Rates, Kar Way Tan, Qi You Ng, Francis Ngoc Hoang Long Nguyen, Sean Shao Wei Lam
Research Collection School Of Computing and Information Systems
Managing bed utilization and ensuring the supply keeps up with the demand is not an easy task in a large public hospital with many medical disciplines. The bed managers who makes decisions on reserving and allocating beds centrally require high-dimensional data from several hospital information systems supporting emergency room, specialized clinics and bed management processes. In this work, we put together an automated process for cleaning, consolidating and integrating data from several hospital information systems to several reports required by the bed managers to analyse the bed occupancy situations across more than thirty medical disciplines. To prevent bed crunch situations …
Data-Driven Surgical Duration Prediction Model For Surgery Scheduling: A Case-Study For A Practice-Feasible Model In A Public Hospital, Kar Way Tan, Francis Ngoc Hoang Long Nguyen, Boon Yew Ang, Jerald Gan, Sean Shao Wei Lam
Data-Driven Surgical Duration Prediction Model For Surgery Scheduling: A Case-Study For A Practice-Feasible Model In A Public Hospital, Kar Way Tan, Francis Ngoc Hoang Long Nguyen, Boon Yew Ang, Jerald Gan, Sean Shao Wei Lam
Research Collection School Of Computing and Information Systems
Hospitals have been trying to improve the utilization of operating rooms as it affects patient satisfaction, surgery throughput, revenues and costs. Surgical prediction model which uses post-surgery data often requires high-dimensional data and contains key predictors such as surgical team factors which may not be available during the surgical listing process. Our study considers a two-step data-mining model which provides a practical, feasible and parsimonious surgical duration prediction. Our model first leverages on domain knowledge to provide estimate of the first surgeon rank (a key predicting attribute) which is unavailable during the listing process, then uses this predicted attribute and …
Learning Multiple Maps From Conditional Ordinal Triplets, Duy Dung Le, Hady Wirawan Lauw
Learning Multiple Maps From Conditional Ordinal Triplets, Duy Dung Le, Hady Wirawan Lauw
Research Collection School Of Computing and Information Systems
Ordinal embedding seeks a low-dimensional representation of objects based on relative comparisons of their similarities. This low-dimensional representation lends itself to visualization on a Euclidean map. Classical assumptions admit only one valid aspect of similarity. However, there are increasing scenarios involving ordinal comparisons that inherently reflect multiple aspects of similarity, which would be better represented by multiple maps. We formulate this problem as conditional ordinal embedding, which learns a distinct low-dimensional representation conditioned on each aspect, yet allows collaboration across aspects via a shared representation. Our geometric approach is novel in its use of a shared spherical representation and multiple …
Sar: Learning Cross-Language Api Mappings With Little Knowledge, Duy Quoc Nghi Bui, Yijun Yu, Lingxiao Jiang
Sar: Learning Cross-Language Api Mappings With Little Knowledge, Duy Quoc Nghi Bui, Yijun Yu, Lingxiao Jiang
Research Collection School Of Computing and Information Systems
To save effort, developers often translate programs from one programming language to another, instead of implementing it from scratch. Translating application program interfaces (APIs) used in one language to functionally equivalent ones available in another language is an important aspect of program translation. Existing approaches facilitate the translation by automatically identifying the API mappings across programming languages. However, these approaches still require large amount of parallel corpora, ranging from pairs of APIs or code fragments that are functionally equivalent, to similar code comments. To minimize the need of parallel corpora, this paper aims at an automated approach that can map …
Multiagent Decision Making And Learning In Urban Environments, Akshat Kumar
Multiagent Decision Making And Learning In Urban Environments, Akshat Kumar
Research Collection School Of Computing and Information Systems
Our increasingly interconnected urban environments provide several opportunities to deploy intelligent agents—from self-driving cars, ships to aerial drones—that promise to radically improve productivity and safety. Achieving coordination among agents in such urban settings presents several algorithmic challenges—ability to scale to thousands of agents, addressing uncertainty, and partial observability in the environment. In addition, accurate domain models need to be learned from data that is often noisy and available only at an aggregate level. In this paper, I will overview some of our recent contributions towards developing planning and reinforcement learning strategies to address several such challenges present in largescale urban …
Trust Architecture And Reputation Evaluation For Internet Of Things, Juan Chen, Zhihong Tian, Xiang Cui, Lihua Yin, Xianzhi Wang
Trust Architecture And Reputation Evaluation For Internet Of Things, Juan Chen, Zhihong Tian, Xiang Cui, Lihua Yin, Xianzhi Wang
Research Collection School Of Computing and Information Systems
Internet of Things (IoT) represents a fundamental infrastructure and set of techniques that support innovative services in various application domains. Trust management plays an important role in enabling the reliable data collection and mining, context-awareness, and enhanced user security in the IoT. The main tasks of trust management include trust architecture design and reputation evaluation. However, existing trust architectures and reputation evaluation solutions cannot be directly applied to the IoT, due to the large number of physical entities, the limited computation ability of physical entities, and the highly dynamic nature of the network. In comparison, it generally requires a general …
A Secure Iot Cloud Storage System With Fine-Grained Access Control And Decryption Key Exposure Resistance, Shengmin Xu, Guomin Yang, Yi Mu, Ximeng Liu
A Secure Iot Cloud Storage System With Fine-Grained Access Control And Decryption Key Exposure Resistance, Shengmin Xu, Guomin Yang, Yi Mu, Ximeng Liu
Research Collection School Of Computing and Information Systems
Internet of Things (IoT) cloud provides a practical and scalable solution to accommodate the data management in large-scale IoT systems by migrating the data storage and management tasks to cloud service providers (CSPs). However, there also exist many data security and privacy issues that must be well addressed in order to allow the wide adoption of the approach. To protect data confidentiality, attribute-based cryptosystems have been proposed to provide fine-grained access control over encrypted data in loT cloud. Unfortunately, the existing attributed-based solutions are still insufficient in addressing some challenging security problems, especially when dealing with compromised or leaked user …
Kgat: Knowledge Graph Attention Network For Recommendation, Xiang Wang, Xiangnan He, Yixin Cao, Meng Liu, Tat-Seng Chua
Kgat: Knowledge Graph Attention Network For Recommendation, Xiang Wang, Xiangnan He, Yixin Cao, Meng Liu, Tat-Seng Chua
Research Collection School Of Computing and Information Systems
To provide more accurate, diverse, and explainable recommendation, it is compulsory to go beyond modeling user-item interactions and take side information into account. Traditional methods like factorization machine (FM) cast it as a supervised learning problem, which assumes each interaction as an independent instance with side information encoded. Due to the overlook of the relations among instances or items (e.g., the director of a movie is also an actor of another movie), these methods are insufficient to distill the collaborative signal from the collective behaviors of users. In this work, we investigate the utility of knowledge graph (KG), which breaks …
Low-Rank Sparse Subspace For Spectral Clustering, Xiaofeng Zhu, Shichao Zhang, Yonggang Li, Jilian Zhang, Lifeng Yang, Yue Fang
Low-Rank Sparse Subspace For Spectral Clustering, Xiaofeng Zhu, Shichao Zhang, Yonggang Li, Jilian Zhang, Lifeng Yang, Yue Fang
Research Collection School Of Computing and Information Systems
The current two-step clustering methods separately learn the similarity matrix and conduct k means clustering. Moreover, the similarity matrix is learnt from the original data, which usually contain noise. As a consequence, these clustering methods cannot achieve good clustering results. To address these issues, this paper proposes a new graph clustering methods (namely Low-rank Sparse Subspace clustering (LSS)) to simultaneously learn the similarity matrix and conduct the clustering from the low-dimensional feature space of the original data. Specifically, the proposed LSS integrates the learning of similarity matrix of the original feature space, the learning of similarity matrix of the low-dimensional …
Who Should Make Decision On This Pull Request? Analyzing Time-Decaying Relationships And File Similarities For Integrator Prediction, Jing Jiang, David Lo, Jiateng Zheng, Xin Xia, Yun Yang, Li Zhang
Who Should Make Decision On This Pull Request? Analyzing Time-Decaying Relationships And File Similarities For Integrator Prediction, Jing Jiang, David Lo, Jiateng Zheng, Xin Xia, Yun Yang, Li Zhang
Research Collection School Of Computing and Information Systems
In pull-based development model, integrators are responsible for making decisions about whether to accept pull requests andintegrate code contributions. Ideally, pull requests are assigned to integrators and evaluated within a short time after their submissions. However, the volume of incoming pull requests is large in popular projects, and integrators often encounter difficulties inprocessing pull requests in a timely fashion. Therefore, an automatic integrator prediction approach is required to assign appropriate pull requests to integrators. In this paper, we propose an approach TRFPre which analyzes Time-decaying Relationships andFile similarities to predict integrators. We evaluate the effectiveness of TRFPre on 24 projects …
Answerbot: An Answer Summary Generation Tool Based On Stack Overflow, Liang Cai, Haoye Wang, Bowen Xu, Qiao Huang, Xin Xia, David Lo, Zhenchang Xing
Answerbot: An Answer Summary Generation Tool Based On Stack Overflow, Liang Cai, Haoye Wang, Bowen Xu, Qiao Huang, Xin Xia, David Lo, Zhenchang Xing
Research Collection School Of Computing and Information Systems
The prevalence of questions and answers on domainspecific Q&A sites like Stack Overflow constitutes a core knowledge asset for software engineering domain. Although search engines can return a list of questions relevant to a user query of some technical question, the abundance of relevant posts and the sheer amount of information in them makes it difficult for developers to digest them and find the most needed answers to their questions. In this work, we aim to help developers who want to quickly capture the key points of several answer posts relevant to a technical question before they read the details …
Correlation-Sensitive Next-Basket Recommendation, Duc Trong Le, Hady Wirawan Lauw, Yuan Fang
Correlation-Sensitive Next-Basket Recommendation, Duc Trong Le, Hady Wirawan Lauw, Yuan Fang
Research Collection School Of Computing and Information Systems
Items adopted by a user over time are indicative ofthe underlying preferences. We are concerned withlearning such preferences from observed sequencesof adoptions for recommendation. As multipleitems are commonly adopted concurrently, e.g., abasket of grocery items or a sitting of media consumption, we deal with a sequence of baskets asinput, and seek to recommend the next basket. Intuitively, a basket tends to contain groups of relateditems that support particular needs. Instead of recommending items independently for the next basket, we hypothesize that incorporating informationon pairwise correlations among items would help toarrive at more coherent basket recommendations.Towards this objective, we develop a …
Knowledge Base Question Answering With Topic Units, Yunshi Lan, Shuohang Wang, Jing Jiang
Knowledge Base Question Answering With Topic Units, Yunshi Lan, Shuohang Wang, Jing Jiang
Research Collection School Of Computing and Information Systems
Knowledge base question answering (KBQA) is an important task in natural language processing. Existing methods for KBQA usually start with entity linking, which considers mostly named entities found in a question as the starting points in the KB to search for answers to the question. However, relying only on entity linking to look for answer candidates may not be sufficient. In this paper, we propose to perform topic unit linking where topic units cover a wider range of units of a KB. We use a generation-and-scoring approach to gradually refine the set of topic units. Furthermore, we use reinforcement learning …
Adversarial Learning On Heterogeneous Information Networks, Binbin Hu, Yuan Fang, Chuan Shi
Adversarial Learning On Heterogeneous Information Networks, Binbin Hu, Yuan Fang, Chuan Shi
Research Collection School Of Computing and Information Systems
Network embedding, which aims to represent network data in alow-dimensional space, has been commonly adopted for analyzingheterogeneous information networks (HIN). Although exiting HINembedding methods have achieved performance improvement tosome extent, they still face a few major weaknesses. Most importantly, they usually adopt negative sampling to randomly selectnodes from the network, and they do not learn the underlying distribution for more robust embedding. Inspired by generative adversarial networks (GAN), we develop a novel framework HeGAN forHIN embedding, which trains both a discriminator and a generatorin a minimax game. Compared to existing HIN embedding methods,our generator would learn the node distribution to …
Multimodal Transformer Networks For End-To-End Video-Grounded Dialogue Systems, Hung Le, Doyen Sahoo, Nancy F. Chen, Steven C. H. Hoi
Multimodal Transformer Networks For End-To-End Video-Grounded Dialogue Systems, Hung Le, Doyen Sahoo, Nancy F. Chen, Steven C. H. Hoi
Research Collection School Of Computing and Information Systems
Developing Video-Grounded Dialogue Systems (VGDS), where a dialogue is conducted based on visual and audio aspects of a given video, is significantly more challenging than traditional image or text-grounded dialogue systems because (1) feature space of videos span across multiple picture frames, making it difficult to obtain semantic information; and (2) a dialogue agent must perceive and process information from different modalities (audio, video, caption, etc.) to obtain a comprehensive understanding. Most existing work is based on RNNs and sequence-to-sequence architectures, which are not very effective for capturing complex long-term dependencies (like in videos). To overcome this, we propose Multimodal …
Creating Top Ranking Options In The Continuous Option And Preference Space, Bo Tang, Kyriakos Mouratidis, Man Lung Yiu, Zhenyu Chen
Creating Top Ranking Options In The Continuous Option And Preference Space, Bo Tang, Kyriakos Mouratidis, Man Lung Yiu, Zhenyu Chen
Research Collection School Of Computing and Information Systems
Top-k queries are extensively used to retrieve the k most relevantoptions (e.g., products, services, accommodation alternatives, etc)based on a weighted scoring function that captures user preferences. In this paper, we take the viewpoint of a business owner whoplans to introduce a new option to the market, with a certain type ofclientele in mind. Given a target region in the consumer spectrum,we determine what attribute values the new option should have,so that it ranks among the top-k for any user in that region. Ourmethodology can also be used to improve an existing option, at theminimum modification cost, so that it ranks …
Adapting Bert For Target-Oriented Multimodal Sentiment Classification, Jianfei Yu, Jing Jiang
Adapting Bert For Target-Oriented Multimodal Sentiment Classification, Jianfei Yu, Jing Jiang
Research Collection School Of Computing and Information Systems
As an important task in Sentiment Analysis, Target-oriented Sentiment Classification (TSC) aims to identify sentiment polarities over each opinion target in a sentence. However, existing approaches to this task primarily rely on the textual content, but ignoring the other increasingly popular multimodal data sources (e.g., images), which can enhance the robustness of these text-based models. Motivated by this observation and inspired by the recently proposed BERT architecture, we study Target-oriented Multimodal Sentiment Classification (TMSC) and propose a multimodal BERT architecture. To model intra-modality dynamics, we first apply BERT to obtain target-sensitive textual representations. We then borrow the idea from self-attention …
Language And Robotics: Complex Sentence Understanding, Seng-Beng Ho, Zhaoxia Wang
Language And Robotics: Complex Sentence Understanding, Seng-Beng Ho, Zhaoxia Wang
Research Collection School Of Computing and Information Systems
Existing robotic systems can take actions based on natural language commands but they tend to be only simple commands. On the other hand, in the domain of Natural Language Processing (NLP), complex sentences are processed, but this NLP domain does not make close contact with robotics. The beginning of computer processing of natural language, when traced back to a system such as Winograd’s SHRUDLU, conceived in 1973, actually aimed to address the issues of Natural Language Understanding (NLU) of relatively complex sentences by a robotic system which in turn takes actions accordingly based on the natural language input. NLU, in …
Ezlog: Data Visualization For Logistics, Aldy Gunawan, Benjamin Gan, Jin An Tan, Sheena L.S.L Villanueva, Timothy K.J. Wen
Ezlog: Data Visualization For Logistics, Aldy Gunawan, Benjamin Gan, Jin An Tan, Sheena L.S.L Villanueva, Timothy K.J. Wen
Research Collection School Of Computing and Information Systems
With the increasing availability of data in the logistics industry due to the digitalization trend, interest and opportunities for leveraging analytics in supply chain management to make data-driven decisions is growing rapidly. In this paper, we introduce EzLog, an integrated visualization prototype platform for supply chain analytics. This web-based platform built by two undergraduate student teams for their capstone course can be used for data wrangling and rapid analysis of data from different business units of a major logistics company. Other functionalities of the system include standard processes to perform data analysis such as supervised extraction, transformation, loading (ETL), data …
Integrated Assignment And Routing With Mixed Service Mode Cross-Dock, Vincent Yu, Aldy Gunawan, Eric I. Junaidi, Audrey T. Widjaja
Integrated Assignment And Routing With Mixed Service Mode Cross-Dock, Vincent Yu, Aldy Gunawan, Eric I. Junaidi, Audrey T. Widjaja
Research Collection School Of Computing and Information Systems
Amixed service mode cross-dock is a cross-dock facility that considers the useof flexible doors. Instead of having a specific task as an exclusive mode, eachdoor can be used as a flexible door, either an inbound or an outbound doordepending on the requirement. Having a mixed service mode cross-dock in anintegrated assignment and routing problem is a new model in large field ofcross-docking problems. Decisions that need to be made include doors’functionality, suppliers’ assignments, customers’ deliveries, and vehicles’ routeswith the objective of minimizing the total transportation and material handlingcosts. We develop a mathematical programming model and propose a SimulatedAnnealing (SA) algorithm …
Biker: A Tool For Bi-Information Source Based Api Method Recommendation, Liang Cai, Haoye Wang, Qiao Huang, Xin Xia, Zhenchang Xing, David Lo
Biker: A Tool For Bi-Information Source Based Api Method Recommendation, Liang Cai, Haoye Wang, Qiao Huang, Xin Xia, Zhenchang Xing, David Lo
Research Collection School Of Computing and Information Systems
No abstract provided.
Inspect: Iterated Local Search For Solving Path Conditions, Fuxiang Chen, Aldy Gunawan, David Lo, Sunghun Kim
Inspect: Iterated Local Search For Solving Path Conditions, Fuxiang Chen, Aldy Gunawan, David Lo, Sunghun Kim
Research Collection School Of Computing and Information Systems
Automated test case generation is attractive as it can reduce developer workload. To generate test cases, many Symbolic Execution approaches first produce Path Conditions (PCs), a set of constraints, and pass them to a Satisfiability Modulo Theories (SMT) solver. Despite numerous prior studies, automated test case generation by Symbolic Execution is still slow, partly due to SMT solvers’ high computationally complexity. We introduce InSPeCT, a Path Condition solver, that leverages elements of ILS (Iterated Local Search) and Tabu List. ILS is not computational intensive and focuses on generating solutions in search spaces while Tabu List prevents the use of previously …
How Does Machine Learning Change Software Development Practices?, Zhiyuan Wan, Xin Xia, David Lo, Gail C. Murphy
How Does Machine Learning Change Software Development Practices?, Zhiyuan Wan, Xin Xia, David Lo, Gail C. Murphy
Research Collection School Of Computing and Information Systems
Adding an ability for a system to learn inherently adds uncertainty into the system. Given the rising popularity of incorporating machine learning into systems, we wondered how the addition alters software development practices. We performed a mixture of qualitative and quantitative studies with 14 interviewees and 342 survey respondents from 26 countries across four continents to elicit significant differences between the development of machine learning systems and the development of non-machine-learning systems. Our study uncovers significant differences in various aspects of software engineering (e.g., requirements, design, testing, and process) and work characteristics (e.g., skill variety, problem solving and task identity). …
Evaluating Vulnerability To Fake News In Social Networks: A Community Health Assessment Model, Bhavtosh Rath, Wei Gao, Jaideep Srivastava
Evaluating Vulnerability To Fake News In Social Networks: A Community Health Assessment Model, Bhavtosh Rath, Wei Gao, Jaideep Srivastava
Research Collection School Of Computing and Information Systems
Understanding the spread of false information in social networks has gained a lot of recent attention. In this paper, we explore the role community structures play in determining how people get exposed to fake news. Inspired by approaches in epidemiology, we propose a novel Community Health Assessment model, whose goal is to understand the vulnerability of communities to fake news spread. We define the concepts of neighbor, boundary and core nodes of a community and propose appropriate metrics to quantify the vulnerability of nodes (individual-level) and communities (group-level) to spreading fake news. We evaluate our model on communities identified using …
A Survey On Bluetooth 5.0 And Mesh: New Milestones Of Iot, Juenjie Yin, Zheng Yang, Hao Cao, Tongtong Liu, Zimu Zhou, Chenshu Wu
A Survey On Bluetooth 5.0 And Mesh: New Milestones Of Iot, Juenjie Yin, Zheng Yang, Hao Cao, Tongtong Liu, Zimu Zhou, Chenshu Wu
Research Collection School Of Computing and Information Systems
No abstract provided.
Cold-Start Aware Deep Memory Networks For Multi-Entity Aspect-Based Sentiment Analysis, Kaisong Song, Wei Gao, Lujun Zhao, Changlong Sun, Xiaozhong Liu
Cold-Start Aware Deep Memory Networks For Multi-Entity Aspect-Based Sentiment Analysis, Kaisong Song, Wei Gao, Lujun Zhao, Changlong Sun, Xiaozhong Liu
Research Collection School Of Computing and Information Systems
Various types of target information have been considered in aspect-based sentiment analysis, such as entities and aspects. Existing research has realized the importance of targets and developed methods with the goal of precisely modeling their contexts via generating target-specific representations. However, all these methods ignore that these representations cannot be learned well due to the lack of sufficient human-annotated target-related reviews, which leads to the data sparsity challenge, a.k.a. cold-start problem here. In this paper, we focus on a more general multiple entity aspect-based sentiment analysis (ME-ABSA) task which aims at identifying the sentiment polarity of different aspects of multiple …
Enhancing Multi-Hop Sensor Calibration With Uncertainty Estimates, Balz Maag, Zimu Zhou, Lothar Thiele
Enhancing Multi-Hop Sensor Calibration With Uncertainty Estimates, Balz Maag, Zimu Zhou, Lothar Thiele
Research Collection School Of Computing and Information Systems
Low-cost sensors, installed on mobile vehicles, provide a cost-effective way for fine-grained urban air pollution monitoring. However, frequent calibration is crucial for lowcost sensors to consistently deliver accurate measurements. Multi-hop calibration is a common practice to calibrate mobile sensor deployments, but is prone to severe error accumulation over hops. Prior research mitigates error accumulation by designing special calibration models, which only apply to linear models. In this paper, we propose an orthogonal approach by selecting reliable measurements for calibration at each hop. We analyze the impact of different data-induced uncertainties on calibration errors and devise a scheme to estimate these …
Faster First-Order Methods For Stochastic Non-Convex Optimization On Riemannian Manifolds, Pan Zhou, Xiao-Tong Yuan, Shuicheng Yan, Jiashi Feng
Faster First-Order Methods For Stochastic Non-Convex Optimization On Riemannian Manifolds, Pan Zhou, Xiao-Tong Yuan, Shuicheng Yan, Jiashi Feng
Research Collection School Of Computing and Information Systems
First-order non-convex Riemannian optimization algorithms have gained recent popularity in structured machine learning problems including principal component analysis and low-rank matrix completion. The current paper presents an efficient Riemannian Stochastic Path Integrated Differential EstimatoR (R-SPIDER) algorithm to solve the finite-sum and online Riemannian non-convex minimization problems. At the core of R-SPIDER is a recursive semi-stochastic gradient estimator that can accurately estimate Riemannian gradient under not only exponential mapping and parallel transport, but also general retraction and vector transport operations. Compared with prior Riemannian algorithms, such a recursive gradient estimation mechanism endows R-SPIDER with higher computational efficiency in first-order oracle complexity. …
Generalized Majorization-Minimization For Non-Convex Optimization, Hu Zhang, Pan Zhou, Yi Yang, Jiashi Feng
Generalized Majorization-Minimization For Non-Convex Optimization, Hu Zhang, Pan Zhou, Yi Yang, Jiashi Feng
Research Collection School Of Computing and Information Systems
Majorization-Minimization (MM) algorithms optimize an objective function by iteratively minimizing its majorizing surrogate and offer attractively fast convergence rate for convex problems. However, their convergence behaviors for non-convex problems remain unclear. In this paper, we propose a novel MM surrogate function from strictly upper bounding the objective to bounding the objective in expectation. With this generalized surrogate conception, we develop a new optimization algorithm, termed SPI-MM, that leverages the recent proposed SPIDER for more efficient non-convex optimization. We prove that for finite-sum problems, the SPI-MM algorithm converges to an stationary point within deterministic and lower stochastic gradient complexity. To our …