Evaluation of clustering and topic modeling methods over health-related tweets and emails

Keyphrases

• Tweets 100%
• Latent Dirichlet Allocation 100%
• Topic Modeling 100%
• Health-related 100%
• Modeling Method 100%
• Topic Clustering 100%
• Gibbs Sampling 83%
• Twitter 66%
• Dirichlet multinomial Mixture 66%
• Clustering Algorithm 50%
• Term Frequency-inverse Document Frequency (TF-IDF) 50%
• K-means 50%
• Clustering Methods 50%
• External Measures 50%
• Latent Semantic Indexing 50%
• Internal Index 50%
• Cluster Validity Index 33%
• Topic Modeling Algorithm 33%
• Model Performance 33%
• Cluster Analysis 16%
• Social Media 16%
• World Wide Web 16%
• Clustering Approach 16%
• Healthcare 16%
• Suboptimal Performance 16%
• Doc2Vec 16%
• Free-text Data 16%
• Feature Representation 16%
• Hamming Loss 16%
• Biterm 16%
• Document Clustering 16%
• Patient Experience 16%
• Number of Subjects 16%
• Number of Clusters 16%
• Loss Metric 16%
• Lynch Syndrome 16%
• Hyperparameters 16%
• Text Dataset 16%
• Data Parameters 16%
• Number of Iterations 16%
• Patient-provider Communication 16%
• Clustering Structure 16%
• Class Label 16%
• Clustering Distribution 16%
• External Information 16%
• Text Data 16%
• Communicating with Patients 16%
• New Data Sources 16%
• Error Analysis 16%

Computer Science

• Gibbs Free Energy 62%