Contents
A Network-Based Approach to Understanding International Cooperation in Environmental Protection
1 Introduction
2 Concepts and Methodology
2.1 Network Data
2.2 Research Design and Methods
3 Research Findings
3.1 One-Mode Analysis of Environmental Agreements and Participating Parties
3.2 Two-Mode Analysis of Environmental Agreements and Participating Parties
4 Discussion and Future Work
References
Critical Mass and Data Integrity Diagnostics of a Twitter Social Contagion Monitor
1 Introduction
2 Background
2.1 Social Movements and Contagion Models
2.2 Social Movements and Social Media
2.3 Social Media Monitoring Tools
3 Methods and Data
3.1 IRB Compliance
3.2 Risks of Collecting Data on Social Movements
3.3 Data Ingestion
3.3.1 Streaming Contagion Monitor
3.3.2 Regional Contagion Monitor
3.3.3 Productionalized Streaming Contagion Monitor
3.4 Candidate Hashtag Selection
3.4.1 Streaming Contagion Monitor
3.4.2 Regional Contagion Monitor
3.4.3 Productionalized Contagion MonitorTM
3.5 Contagion Analysis
3.5.1 Streaming Contagion Monitor
3.5.2 Regional Contagion Monitor
3.5.3 Productionalized Contagion Monitor
3.6 Reporting
3.7 Coordination Framework
3.7.1 Introduction
3.7.2 Implementation: Data Ingestion
3.7.3 Implementation: Outputs
3.7.4 Integration
4 Results
4.1 Framework for Evaluating the Contagion Monitors
4.2 Evaluation of Complexity of Contagion (R1)
4.2.1 Mechanical Turk Annotation
4.2.2 Analysis of Annotated Hashtags
4.2.3 Spam Filtering
4.2.4 Linear Regression
4.2.5 Classification
4.3 Analysis of Critical Mass for Virality (R2)
4.4 Analysis of Data Size (R3)
4.5 A Fruitful Case Study
4.6 Results with Coordination Framework Metrics
5 Discussion and Conclusion
Appendix
References
TenFor: Tool to Mine Interesting Events from Security Forums Leveraging Tensor Decomposition
1 Introduction
2 Background and Datasets
2.1 Datasets
2.2 Background
3 Our Approach
3.1 Step 1: Tensor-Based Clustering
3.2 Step 2: Profiling the Clusters
3.3 Step 3: Investigation of Clusters
4 Results and Evaluation
4.1 Step by Step Output Provided by TenFor
4.2 Evaluation of TenFor with Real Data
4.3 Evaluation of TenFor with Synthetic Data
5 Discussion
6 Related Work
7 Conclusion
References
Profile Fusion in Social Networks: A Data-Driven Approach
1 Introduction
2 Related Works
3 ULSN System Architecture
3.1 Re-Usability, Code and Implementation Details
4 Data
4.1 Data Validation
4.2 Dataset Overview
4.3 Content Analysis
4.4 Temporal Analysis
5 User Profile Linkage
5.1 Problem Formulation
5.2 User Matching Approach
5.3 Classification Approach
5.4 Clustering-Based Optimization
6 Experiments
6.1 User Matching Results
6.2 Classification Results
6.3 The Impact of Clustering
6.4 The Impact of User Prolificacy
6.5 The Impact of Time
7 Perspectives on Exploiting ULSN Data
7.1 User Matching
7.2 Expertise Retrieval
7.3 Authorship Identification
8 Conclusion
References
RISECURE: Metro Transit Disruptions Detection Using Social Media Mining And Graph Convolution
1 Introduction
2 Related Work
3 System Overview
3.1 Data Acquisition
3.2 Application Server
3.2.1 Application and Mobile Interface
3.2.2 Real Time Incidents Panel
3.2.3 Alert Notification System
3.2.4 Station Marker
4 Methodology
4.1 Dynamic Query Expansion
4.2 Graph Convolution
4.2.1 Building The Graph
4.2.2 Network Architecture
5 Experiment and Results
5.1 Benchmark Evaluations
5.2 Parameter Testing
5.2.1 Size of Sliding Window
5.2.2 Size of Embedding Dimension
5.2.3 Size of Training Data
6 Conclusion
References
Local Taxonomy Construction: An Information Retrieval Approach Using Representation Learning
1 Introduction
2 Related Work
3 Local Taxonomy Construction
3.1 Framing LTC as an Information Retrieval (IR) Problem Instance
3.2 Evaluating IR Results
4 Solutions
4.1 Representation Learning Approaches for LTC
5 Experiments
5.1 Data
5.2 Methods and Parameters
5.3 Methodology and Metrics
5.4 Results
6 Discussion
7 Potential Applications
8 Future Work
9 Conclusion
References
The Evolution of Online Sentiments Across Italy During First and Second Wave of the COVID-19 Pandemic
1 Introduction
2 Related Literature
3 Data Collection and Description
3.1 Twitter
3.2 Socio-Economic and Epidemiological Variables
4 Methodology
4.1 Extracting the Online Sentiment
4.2 Regression Model
5 Results and Discussion
5.1 The First Wave of Contagion
5.2 The Second Wave of Contagion and Main Differences with the Previous Period
6 Conclusion
References
Inferring Degree of Localization and Popularity of Twitter Topics and Persons Using Temporal Features
1 Introduction
2 Related Research
3 UTC Offset Prediction Based on Account Creation
3.1 UTC Offset Dataset
3.2 Sleep Cycle and UTC Offset Determination
3.3 Parameter Determination
4 Temporal Analysis of Message Traffic Data
4.1 Message Traffic Dataset
4.2 Predicting Region of Token
4.3 Evaluation
4.4 Comparison Against Baseline Based on Google Trends
5 Evolving Popularity: Inferring Daily Changes in Number of Followers
5.1 Dataset: Stable, Global, Growing Influencers
5.1.1 Data Collected for Each Influencer
5.2 An Algorithm to Estimate Follower Gain
5.2.1 Representing Cyclical Nature of Time
5.2.2 The Algorithm
5.3 Evaluation
5.3.1 Baseline 1
5.3.2 Baseline 2
5.4 Rationale for Proposed Algorithm and Its Limitations
5.5 Studying the Evolution of Popularity
6 Global vs. Local Influencer Classifier
6.1 Dataset
6.2 Features
6.3 Results: Local Versus Global Classification
7 Conclusions
References
Covid-19 and Vaccine Tweet Analysis
1 Introduction
2 Literature Review
3 Methodology
3.1 Data Collection
3.2 Preprocessing
3.3 Vectorization
3.4 Sentiment Analysis
3.5 Visaulization
4 Result and Discussion
5 Conclusion
References