0 ×

01_​Clustering_​the_​Social_​Community

Workflow

Clustering Social Media Community

This workflow clusters social media users based on their authority (leader) and hub (follower) score and on their sentiment attitude. The attitude and leader scores have been calculated previously in another workflow (see "Usable Customer Intelligence from Social Media Data: Network Analytics meets Text Mining" whitepaper accessible via the attached link).

social mediaclusteringsentiment analysisk-Means
Read and Normalize Data This workflow performs k-Means clustering on social media users. Metanode "Extract features from K-Means model" is now redundant, since thenew version of the K-Means node already produces the cluster center at thesecond output port.It still is a great example for XML processing in KNIME... ReportingIn order to see the report execute the entire workflow and then click "Open thereport" button in the toolbar. Text+network combinedyou need to enter a valid pathAttitude colorgreen= positivered= negativegray = neutralExtract prototypesfeatures from modelPrototypesUndersampling of all classes till negative attitude class numberStd by clusterAdd std to prototypesCluster_3Cluster_1cluster_1By authority scoreBy authority scorecluster_3Normalizations ofAuth ScoreHub ScoreAttitudeK = 10Authority Scorevs.Hub Score Table Reader Column Filter Color Manager Extracts featuresfrom K-Means model Data to Report Equal Size Sampling GroupBy Joiner Cluster Assigner Row Filter Row Filter Data to Report Sorter Sorter Data to Report Normalizations k-Means Scatter Plot Read and Normalize Data This workflow performs k-Means clustering on social media users. Metanode "Extract features from K-Means model" is now redundant, since thenew version of the K-Means node already produces the cluster center at thesecond output port.It still is a great example for XML processing in KNIME... ReportingIn order to see the report execute the entire workflow and then click "Open thereport" button in the toolbar. Text+network combinedyou need to enter a valid pathAttitude colorgreen= positivered= negativegray = neutralExtract prototypesfeatures from modelPrototypesUndersampling of all classes till negative attitude class numberStd by clusterAdd std to prototypesCluster_3Cluster_1cluster_1By authority scoreBy authority scorecluster_3Normalizations ofAuth ScoreHub ScoreAttitudeK = 10Authority Scorevs.Hub ScoreTable Reader Column Filter Color Manager Extracts featuresfrom K-Means model Data to Report Equal Size Sampling GroupBy Joiner Cluster Assigner Row Filter Row Filter Data to Report Sorter Sorter Data to Report Normalizations k-Means Scatter Plot

Download

Get this workflow from the following link: Download

Resources

Nodes

01_​Clustering_​the_​Social_​Community consists of the following 34 nodes(s):

Plugins

01_​Clustering_​the_​Social_​Community contains nodes provided by the following 7 plugin(s):