Yankun Zhen

ORCID: 0000-0002-7153-1090
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Software Engineering Research
  • Web Data Mining and Analysis
  • Software Engineering Techniques and Practices
  • Video Analysis and Summarization
  • Human Mobility and Location-Based Analysis
  • Mobile and Web Applications
  • Interactive and Immersive Displays
  • Innovative Human-Technology Interaction
  • Virtual Reality Applications and Impacts
  • Open Source Software Innovations
  • Data Visualization and Analytics
  • Augmented Reality Applications
  • Natural Language Processing Techniques
  • Advanced Image and Video Retrieval Techniques
  • Software Testing and Debugging Techniques
  • Multimodal Machine Learning Applications
  • Topic Modeling
  • Advanced Malware Detection Techniques
  • Vehicle License Plate Recognition

Alibaba Group (China)
2021-2024

Approximately 50% of development resources are devoted to UI tasks [8]. Occupied a large proportion resources, developing icons can be time-consuming task, because developers need consider not only effective implementation methods but also easy-to-understand descriptions. In this study, we define 100 icon classes through an iterative open coding for the existing design sharing website. Based on deep learning model and computer vision methods, propose approach automatically convert images...

10.1145/3397481.3450671 article EN 2021-04-14

Code generation aims to generate a code snippet automatically from natural language descriptions. Generally, the mainstream methods rely on large amount of paired training data, including both description and code. However, in some domain-specific scenarios, building such corpus for is difficult because there no directly available pairing lot effort required manually write descriptions construct high-quality dataset. Due limited model cannot be well trained likely overfitting, making model's...

10.1145/3540250.3558965 article EN Proceedings of the 30th ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering 2022-11-07

Approximately 50% of development resources are devoted to UI tasks [9]. Occupying a large proportion resources, developing icons can be time-consuming task, because developers need consider not only effective implementation methods but also easy-to-understand descriptions. In this paper, we present Auto-Icon+, an approach for automatically generating readable and efficient code from design artifacts. According our interviews understand the gap between designers (icons assembled multiple...

10.1145/3531065 article EN ACM Transactions on Interactive Intelligent Systems 2022-04-28

When translating UI design prototypes to code in industry, automatically generating from can expedite the development of applications and GUI iterations. However, without strict specifications, components may be composed fragmented elements. Grouping these elements greatly improve readability maintainability generated code. Current methods employ a two-stage strategy that introduces hand-crafted rules group Unfortunately, performance is not satisfying due visually overlapped tiny In this...

10.1145/3597503.3623313 preprint EN 2024-02-06

Layout is essential for the product listing pages (PLPs) in mobile shopping applications. To clearly convey information that consumers require and to achieve specific functions, PLPs layouts often have many variations driven by scenarios. In this work, we study layout design different scenarios propose a space guide large-scale creation of PLPs. We LayoutVQ-VAE, novel model specialized generating with internal external constraints. LayoutVQ-VAE differs from previous methods as it learns...

10.1145/3544548.3581446 article EN 2023-04-19

Texts, widgets, and images on a UI page do not work separately. Instead, they are partitioned into groups to achieve certain interaction functions or visual information. Existing studies elements grouping mainly focus specific single UI-related software engineering task, their vary in appearance function. In this case, we propose our semantic component that pack adjacent text non-text with similar semantics. contrast those task-oriented methods, group can be adopted for multiple tasks, such...

10.48550/arxiv.2403.04984 preprint EN arXiv (Cornell University) 2024-03-07

Abstract Traditional VR shopping applications often rely on static designs for spaces, necessitating frequent teleportation. However, an exploratory case study involving experienced online shoppers revealed several issues with this approach, including problems interactivity, efficiency, motion sickness, and product presentation. To address these challenges, a workshop was organized professional designers to gather design strategies, proposing the concept of dynamic rearrangement space. This...

10.1115/detc2024-141136 article EN 2024-08-25
Coming Soon ...