A5103096676- Data Management and Algorithms
- Recommender Systems and Techniques
- Advanced Database Systems and Queries
- Advanced Bandit Algorithms Research
- Consumer Market Behavior and Pricing
- Human Mobility and Location-Based Analysis
- Advanced Image and Video Retrieval Techniques
- Time Series Analysis and Forecasting
- Web Data Mining and Analysis
- Computational Geometry and Mesh Generation
- Automated Road and Building Extraction
- Geographic Information Systems Studies
- Advanced Memory and Neural Computing
- Brain Tumor Detection and Classification
- IoT and Edge/Fog Computing
- Data Mining Algorithms and Applications
- Data Quality and Management
- Algorithms and Data Compression
- Optimization and Search Problems
- Smart Parking Systems Research
University of Hong Kong
2013-2020
Hong Kong University of Science and Technology
2015
Recommending packages of items to groups users has several applications, including recommending vacation tourists, entertainment friends, or sets courses students. In this paper, we focus on a novel aspect package-to-group recommendations, that fairness. Specifically, when recommend package group people, ask recommendation is fair in the sense every member satisfied by sufficient number package. We explore two definitions fairness and show for either definition problem finding most NP-hard....
The success of recommender systems has made them the focus a massive research effort in both industry and academia. Recent work generalized recommendations to suggest packages items single users, or groups users. However, best our knowledge, interesting problem recommending package group users (P2G) not been studied date. This is with several practical applications, such as vacation tourist groups, entertainment friends, sets courses students. In this paper, we formulate P2G problem, propose...
Keyword suggestion in web search helps users to access relevant information without having know how precisely express their queries. Existing keyword techniques do not consider the locations of and query results; i.e., spatial proximity a user retrieved results is taken as factor recommendation. However, relevance many applications (e.g., location-based services) known be correlated with issuer. In this paper, we design location-aware framework. We propose weighted keyword-document graph,...
Consider a user who has issued keyword query to search engine. We study the effective suggestion of alternative queries user, which are semantically relevant original and they have as results documents that correspond objects near user's location. For this purpose, we propose weighted keyword-document graph captures semantic proximity relevance between documents. Then, use suggest in terms distance queries. To make our framework scalable, partition-based approach greatly outperforms baseline...
Top-k joins have been extensively studied in relational databases as ranking operations when every object has, among others, at least one attribute. However, the focus has mostly case join attributes are of primitive data types (e.g., numerical values) and predicate is equality. In this work, we consider string objects assigned such or simply scores. Given two collection a similarity measure Edit distance), introduce top-k () which returns k sufficiently similar pairs with respect to...
Distributed machine learning (DML) technology makes it possible to train large neural networks in a reasonable amount of time. Meanwhile, as the computing power grows much faster than network capacity, communication has gradually become bottleneck DML. Current multi-tenant GPU clusters face contention caused by hash-collision problem which not only further increases overhead communication, but also creates unfairness and affects user experience. In this paper, we firstly analyse how training...