Search

Your search keyword '"Tan, Yiming"' showing total 121 results
Start Over Author "Tan, Yiming"
121 results on '"Tan, Yiming"'

1. Incorporating Domain Knowledge Graph into Multimodal Movie Genre Classification with Self-Supervised Attention and Contrastive Learning

Author

Li, Jiaqi, Qi, Guilin, Zhang, Chuanyi, Chen, Yongrui, Tan, Yiming, Xia, Chenlong, and Tian, Ye

Subjects
Computer Science - Artificial Intelligence
Abstract

Multimodal movie genre classification has always been regarded as a demanding multi-label classification task due to the diversity of multimodal data such as posters, plot summaries, trailers and metadata. Although existing works have made great progress in modeling and combining each modality, they still face three issues: 1) unutilized group relations in metadata, 2) unreliable attention allocation, and 3) indiscriminative fused features. Given that the knowledge graph has been proven to contain rich information, we present a novel framework that exploits the knowledge graph from various perspectives to address the above problems. As a preparation, the metadata is processed into a domain knowledge graph. A translate model for knowledge graph embedding is adopted to capture the relations between entities. Firstly we retrieve the relevant embedding from the knowledge graph by utilizing group relations in metadata and then integrate it with other modalities. Next, we introduce an Attention Teacher module for reliable attention allocation based on self-supervised learning. It learns the distribution of the knowledge graph and produces rational attention weights. Finally, a Genre-Centroid Anchored Contrastive Learning module is proposed to strengthen the discriminative ability of fused features. The embedding space of anchors is initialized from the genre entities in the knowledge graph. To verify the effectiveness of our framework, we collect a larger and more challenging dataset named MM-IMDb 2.0 compared with the MM-IMDb dataset. The experimental results on two datasets demonstrate that our model is superior to the state-of-the-art methods. We will release the code in the near future., Comment: Accepted by ACM MM 2023

Published
2023

2. Characteristics and Influencing Factors of Floating Population's Settlement Intention in the Pearl River Delta

Author

Li Zhiyan, Tan Yiming, and Huang Gengzhi

Subjects
floating population, settlement intention, migration, multiple logistic regression, pearl river delta, Geography (General), G1-922
Abstract

During the stage of high-quality development, the orderly promotion of citizenization of the floating population is crucial to ensure the high-quality population development and sustainable socio-economic growth of our country. In this study, dynamic monitoring survey data of China's floating population in 2017 were referenced to categorize the settlement intentions of this population group into three types (viz., short term, long term, and permanent) according to their expected residence duration. Subsequently, the mobility and residential characteristics of this population were analyzed by using a multinomial logistic regression model to examine differences in their demographic and family characteristics, mobility patterns, economic status, and social integration. The findings are as follows: (1) Dongguan and Huizhou, located on the east bank of the Pearl River and primarily driven by manufacturing industries, have a significant proportion of populations with short-term settlement intentions. By contrast, cities with lower living costs or more comprehensive services are more likely to attract long-term residents. (2) The inter-provincial floating population tends to flow into Shenzhen, with individuals migrating a short distance being more predominant. Long-distance inter-provincial migrants have a stronger intention to settle permanently. Within the province, the floating population tends to flow into Guangzhou, displaying a certain degree of geographical proximity in city choices and a tendency to stay long term or permanently. (3) Groups characterized by female sex, older age, higher educational level, marital status, longer inflow time, non-work-related inflow reasons, and higher income level tend to take up long-term or permanent residence at their destination. Conversely, lower levels of these characteristics are associated with shorter stays. Higher degrees of social integration can promote permanent settlement intentions among inter-provincial migrants compared with the effect on their intra-provincial counterparts. (4) Each factor has different effects on groups with the same settlement intention within and outside the province, as reflected in the varying impacts of educational level, marital status, income level, and migration reasons on inter- and intra-provincial floating populations. This study breaks through the traditional binary division model of settlement intentions, enhancing our understanding of floating populations through the "short-term residence–long-term residence–permanent settlement" categorization method. The Pearl River Delta region, one of the areas with the largest concentration of the floating population and the highest degree of economic development in China, serves as a critical case study. Understanding the flow and residence patterns of this population in this region will help government departments implement relevant policies and safeguarding measures tailored to different settlement intentions, thereby improving urban operational efficiency.

Published
2024
Full Text
View/download PDF

3. Can ChatGPT Replace Traditional KBQA Models? An In-depth Analysis of the Question Answering Performance of the GPT LLM Family

Author

Tan, Yiming, Min, Dehai, Li, Yu, Li, Wenbo, Hu, Nan, Chen, Yongrui, and Qi, Guilin

Subjects
Computer Science - Computation and Language
Abstract

ChatGPT is a powerful large language model (LLM) that covers knowledge resources such as Wikipedia and supports natural language question answering using its own knowledge. Therefore, there is growing interest in exploring whether ChatGPT can replace traditional knowledge-based question answering (KBQA) models. Although there have been some works analyzing the question answering performance of ChatGPT, there is still a lack of large-scale, comprehensive testing of various types of complex questions to analyze the limitations of the model. In this paper, we present a framework that follows the black-box testing specifications of CheckList proposed by Ribeiro et. al. We evaluate ChatGPT and its family of LLMs on eight real-world KB-based complex question answering datasets, which include six English datasets and two multilingual datasets. The total number of test cases is approximately 190,000. In addition to the GPT family of LLMs, we also evaluate the well-known FLAN-T5 to identify commonalities between the GPT family and other LLMs. The dataset and code are available at https://github.com/tan92hl/Complex-Question-Answering-Evaluation-of-GPT-family.git, Comment: To be published in Proceedings of ISWC 2023, 22nd International Semantic Web Conference

Published
2023

12. MAKG: A Mobile Application Knowledge Graph for the Research of Cybersecurity

Author

Zhou, Heng, Li, Weizhuo, Zhang, Buye, Ji, Qiu, Tan, Yiming, Na, Chongning, Filipe, Joaquim, Editorial Board Member, Ghosh, Ashish, Editorial Board Member, Prates, Raquel Oliveira, Editorial Board Member, Zhou, Lizhu, Editorial Board Member, Qin, Bing, editor, Jin, Zhi, editor, Wang, Haofen, editor, Pan, Jeff, editor, Liu, Yongbin, editor, and An, Bo, editor

Published
2021
Full Text
View/download PDF

19. Automatic Generation of Multiple-Choice Items for Prepositions Based on Word2vec

Author

Xiao, Wenyan, Wang, Mingwen, Zhang, Chenlin, Tan, Yiming, Chen, Zhiming, Barbosa, Simone Diniz Junqueira, Series Editor, Filipe, Joaquim, Series Editor, Kotenko, Igor, Series Editor, Sivalingam, Krishna M., Series Editor, Washio, Takashi, Series Editor, Yuan, Junsong, Series Editor, Zhou, Lizhu, Series Editor, Zhou, Qinglei, editor, Miao, Qiguang, editor, Wang, Hongzhi, editor, Xie, Wei, editor, Wang, Yan, editor, and Lu, Zeguang, editor

Published
2018
Full Text
View/download PDF

24. Evaluation of ChatGPT as a Question Answering System for Answering Complex Questions

Author

Tan, Yiming, Min, Dehai, Li, Yu, Li, Wenbo, Hu, Nan, Chen, Yongrui, and Qi, Guilin

Subjects
FOS: Computer and information sciences, Computer Science - Computation and Language, Computation and Language (cs.CL)
Abstract

ChatGPT is a powerful large language model (LLM) that has made remarkable progress in natural language understanding. Nevertheless, the performance and limitations of the model still need to be extensively evaluated. As ChatGPT covers resources such as Wikipedia and supports natural language question answering, it has garnered attention as a potential replacement for traditional knowledge based question answering (KBQA) models. Complex question answering is a challenge task of KBQA, which comprehensively tests the ability of models in semantic parsing and reasoning. To assess the performance of ChatGPT as a question answering system (QAS) using its own knowledge, we present a framework that evaluates its ability to answer complex questions. Our approach involves categorizing the potential features of complex questions and describing each test question with multiple labels to identify combinatorial reasoning. Following the black-box testing specifications of CheckList proposed by Ribeiro et.al, we develop an evaluation method to measure the functionality and reliability of ChatGPT in reasoning for answering complex questions. We use the proposed framework to evaluate the performance of ChatGPT in question answering on 8 real-world KB-based CQA datasets, including 6 English and 2 multilingual datasets, with a total of approximately 190,000 test cases. We compare the evaluation results of ChatGPT, GPT-3.5, GPT-3, and FLAN-T5 to identify common long-term problems in LLMs. The dataset and code are available at https://github.com/tan92hl/Complex-Question-Answering-Evaluation-of-ChatGPT.

Published
2023
Full Text
View/download PDF

35. Quality Control in Construction Process of Highway Mechanical and Electrical System

Author

Zhu Liwei, Tan Yiming, and Tang Junjun

Subjects
Electric power system, Computer science, Service level, media_common.quotation_subject, Control (management), Key (cryptography), Process (computing), Topology (electrical circuits), Quality (business), Function (engineering), media_common, Reliability engineering
Abstract

The quality of Highway Mechanical and Electrical System (HMES) in construction process is considered as the key link to ensure the highway operational safety and service level. In this paper, the definition and characterization of the quality of HMES in construction process is given, the functional topology structure of HMES is provided, and subsequently the quality transfer model based on the function topology of HMES is presented. In the end of this paper, the experiment result of a given Highway Tunnel Mechanical and Electrical System (HTMES) is provided to verify the effectiveness of the proposed quality transfer model.

Published
2020

40. Thermal Conductivity of 2CaO·SiO2 Bearing Solid Solution

Author

Rie Endo, Tan Yiming, Yoshinao Kobayashi, and Shunsuke Takahashi

Subjects
010302 applied physics, Bearing (mechanical), Materials science, Mechanical Engineering, Metallurgy, Metals and Alloys, Mineralogy, 02 engineering and technology, engineering.material, 01 natural sciences, 020501 mining & metallurgy, law.invention, Thermal conductivity, 0205 materials engineering, Mechanics of Materials, law, 0103 physical sciences, Materials Chemistry, engineering, Dissolution, Solid solution, Lime
Published
2017

42. Elasmoid fish scales as a natural fibre composite: microscopic heterogeneities in structure, mineral distribution, and mechanical properties

Author

Tan, Yiming, Jia, Zian, Deng, Zhifei, and Li, Ling

Abstract

The elasmoid scales in teleost fish serve as exemplary models for natural fibre composites with integrated flexibility and protection. Yet, limited research has been focused on the potential structural, chemical, and mechanical heterogeneity within individual scales. This study presents systematic characterizations of the elasmoid scales from black drum fish (Pogonias cromis) at different zones within individual scales as a natural fibre composite, focusing on the microscopic structural heterogeneities and corresponding mechanical effects. The focus field at the centre of the scales exhibits a classical tri-layered collagen-based composite design, consisting of the mineralized outermost limiting layer, external elasmodine layer in the middle, and the unmineralized internal elasmodine layer. In comparison, the rostral field at the anterior end of the scales exhibits a two-layered design: the mineralized outermost limiting layer exhibits radii sections on the outer surface, and the inner elasmodine layer consists of collagen fibre-based sublayers with alternating mineralization levels. Chemical and nanoindentation analysis suggests a close correlation between the mineralization levels and the local nanomechanical properties. Comparative finite element modelling shows that the rostral-field scales achieve increased flexibility under both concave and convex bending. Moreover, the evolving geometries of isolated Mandle’s corpuscles in the internal elasmodine layer, transitioning from irregular shapes to faceted octahedrons, suggest the mechanisms of mineral growth and space-filling to thicken the mineralized layers in scales during growth, which enhances the bonding strength between the adjacent collagen fibre layers. This work offers new insights into the structural variations in individual elasmoid scales, providing strategies for bioinspired fibre composite designs with local-adapted functional requirements.

Published
2024
Full Text
View/download PDF

43. Discussion of active safety countermeasures of motorway work zone

Author

Sun Xiaohong, Hongxi Yu, Zhang Zhengyu, Tan Yiming, Chai Hua, and Junjun Tang

Subjects
Work zone, Active safety, Forensic engineering, Environmental science
Abstract

A series of comparison analyses of work zone configurations, traffic organizations and work management among China and oversea countries are carried out. Taking into consideration of human, vehicles, roads, environment as well as management factors, the causes of traffic accidents in work zone areas are deeply discussed. Based on road safety in work zone areas, a series of active traffic safety improvement countermeasures for both workers and drivers in work zone areas have been introduced from the respects from traditional safety devices and traffic control measures to information and intelligent measures, which will provide inspires for the improvement of work safety and operation safety in work zone areas.

Published
2019

47. Division by countries of origin? Visualizing the spatial distributions of international migrants in Guangzhou, China

Author

Tan, Yiming and Chen, Zifeng

Abstract

Socio-spatial segregation of immigrants or other ethno-racial groups in Western cities has been extensively investigated. In the recent decades, China has also witnessed a substantial growth of international immigrants. In the city of Guangzhou, one of the most famous destinations in China for transnational migration, the spatial presence of international migrants has received scholarly attention, mostly focusing on single racial groups. In this study, we present two cartograms using cellphone data to visualize the spatial distributions of multiple groups of international migrants, namely, the African migrants, the European and North American migrants, and the Japanese and Korean migrants, in Guangzhou. The cartograms indicate that the spatial distributions of migrants from Africa and those from the European, North American, and East Asian countries are considerably divided in Guangzhou, suggesting a possible ethno-racial segregation among the international migrants in this Chinese city. Such an issue is largely under-researched in the existing literature.

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results