A5010929778- Complex Network Analysis Techniques
- Opinion Dynamics and Social Influence
- Recommender Systems and Techniques
- Digital Marketing and Social Media
- Advanced Graph Neural Networks
- Human Mobility and Location-Based Analysis
- Caching and Content Delivery
- Sentiment Analysis and Opinion Mining
- Social Media and Politics
- Complex Systems and Time Series Analysis
- Topic Modeling
- Knowledge Management and Sharing
- Evolutionary Game Theory and Cooperation
- Technology Adoption and User Behaviour
- Diverse Aspects of Tourism Research
- International Business and FDI
- scientometrics and bibliometrics research
- Market Dynamics and Volatility
- Mental Health Research Topics
- Firm Innovation and Growth
- Analytical Chemistry and Sensors
- Sleep and Work-Related Fatigue
- Data Quality and Management
- Consumer Behavior in Brand Consumption and Identification
- Health Literacy and Information Accessibility
Nanjing University of Information Science and Technology
2020-2024
Tsinghua University
2023
University of Reading
2016-2022
City University of Hong Kong
2019-2021
Hebei University of Science and Technology
2020
Jilin Medical University
2011-2020
Chinese Academy of Social Sciences
2020
James Cook University
2018
Beijing Jiaotong University
2018
Chinese Academy of Sciences
2009-2018
Abstract Similarity is a fundamental measure in network analyses and machine learning algorithms, with wide applications ranging from personalized recommendation to socio-economic dynamics. We argue that an effective similarity measurement should guarantee the stability even under some information loss. With six bipartite networks, we investigate stabilities of fifteen measurements by comparing matrixes two data samples which are randomly divided original sets. Results show that, can be well...
The mechanism of the online user preference evolution is great significance for understanding behaviors and improving quality services. Since users are allowed to rate on objects in many systems, ratings can well reflect users' preference. With two benchmark datasets from we uncover memory effect selecting behavior which sequence qualities selected rating delivered by each user. Furthermore, duration presented describe length a memory, exhibits power-law distribution, i.e., probability...
The structures of feature vectors-based semisupervised/supervised learning have gained considerable interest in recent years due to their effectiveness for better object modeling and classification. In many machine computer vision tasks, a critical issue is the similarity between two vectors. this paper, we present novel technique measure similarities among vectors by decomposing each vector as an ℓ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML"...
Social media and online navigation bring us enjoyable experiences in accessing information, simultaneously create information cocoons (ICs) which we are unconsciously trapped with limited biased information. We provide a formal definition of IC the scenario navigation. Subsequently, by analyzing real recommendation networks extracted from Science, PNAS, Amazon websites, testing mainstream algorithms disparate recommender systems, demonstrate that similarity-based techniques result ICs,...
Network-based similarity measures have found wide applications in recommendation algorithms and made significant contributions for uncovering users' potential interests. However, existing are generally biased terms of popularity, that the popular objects tend to more common neighbours with others thus considered similar others. Such popularity bias quantification will result recommendations, either poor accuracy or diversity. Based on bipartite network modelling user-object interactions,...
Introduction Due to its effectiveness and various benefits, the use of online health consultation (OHC) has dramatically increased in recent years, especially since outbreak COVID-19 pandemic. However, underlying mechanism whereby pandemic impacted OHC usage is still unclear. Methods Via an survey (N=318), present paper measures users’ perceptions towards both offline services, their intention switch OHC, perceived risks. The relationships among these factors are conceptualized by...
To address the data ‘islandization’ issue in statistical field and to take advantage of opportunity Statistical Cloud construction, National Bureau Statistics China (NBS) started adopting concept a “data middle platform” for resource planning. With it, NBS aims build comprehensive capability platform that includes collection exchange; sharing integration; organizing processing; modeling analyses; management governance; service application. The provides basic application support. It also...
Scaling invariance of time series has been making great contributions in diverse research fields. But how to evaluate scaling exponent from a real-world is still an open problem. Finite length may induce unacceptable fluctuation and bias statistical quantities consequent invalidation currently used standard methods. In this paper new concept called correlation-dependent balanced estimation diffusion entropy developed scale-invariance very short with . Calculations specified Hurst values show...
Despite enormous recent efforts in detecting the mechanism of social relation formation online systems, underlying rules between common interests and relations are still under dispute. Do users befriend others who have similar tastes, or do their tastes become more after they friends? In this paper, we investigate correlation user trust interests, measured by overlap rate ρ taste similarity θ respectively. The creation time is set as zero timestamp. statistical results before for an network,...
The clustering coefficient of the bipartite network, C 4 , has been widely used to investigate statistical properties user-object systems. In this paper, we empirically analyze evolution patterns for a nine year MovieLens data set, where is describe diversity user interest. First, divide set into fractions according time intervals and calculate each fraction. empirical results show that, interest changes periodically with round one year, which reaches smallest value in spring, then increases...
Patient-reported posts in Online Health Communities (OHCs) contain various valuable information that can help establish knowledge-based online support for patients. However, utilizing these reports to improve patient services the absence of appropriate medical and healthcare expert knowledge is difficult. Thus, we propose a comprehensive discovery method based on Unified Medical Language System analysis narrative OHCs. First, domain-knowledge framework OHCs provide basis post analysis....
Purpose User-generated content (UGC) refers to semantic and behavioral traces created by users on various social media platforms. While several waves of platforms have come gone, the long-term sustainability UGC activities has become a critical question that bears significance for theoretical understanding practices. Design/methodology/approach Based large lengthy dataset both blogging microblogging same set users, multistate survival analysis was applied explore patterns users' staying,...
Social influence drives human selection behaviours when numerous objects competing for limited attentions, which leads to the 'rich get richer' dynamics where popular tend more attentions. However, evidences have been found that, both global information of whole system and local among one's friends significant over selection. Consequently, a key question raises it is or determinative selection? Here we compare local-based global-based influence. We show behaviour mainly driven by popularity...