Demo Browser of the existing Visualization Tools.
-
Chord Diagram
Chord Diagram of the co-occurrences of most frequent Named Entities per Time Period. Select 1 of the 8 Time Periods from the dropdown menu. Then, the chord diagram depicts the frequency of co-occurence of the different pairs of Named Entities that were found in the data. The Named Entities are positioned on the diagram, starting from 12 o'clock and going clockwise, relative to their frequency.
-
Network of Users
This is a User-Topic Network, containing a percentage of Users as nodes. The Users were filtered based on their deegre range to make the visualization readable. The Node Radius shows the Mean Influence of User. The Node Color depends on the most proable topic of the user, based on the topic extraction methodology (15 different topics). The Links are directed links without weights.
-
Time Series
Time Series of the Mean Influence of the top-10 Named Entities for each Time Period. Select 1 of the 8 Time Periods from the dropdown menu. Then, you will see how the top-10 most frequent Named Entities, for the selected period, fared over whole time-span in terms of Mean Influence. The Mean Influence of the Named Entities is calculated as the mean of the influence of each user that that mentioned the Named Entity in their tweet.
-
Relation Extraction
The Relation Extraction Module as described. Input a text, tweet or a sentence and extract the relations contained, based on the Named Entities of the input.
-
Named Entities Network
The Named Entities Network contains a percentage of the Named Entities extracted from the tweets as nodes and linked by co-occurence. The Named Entities were selected as the most frequent ones. The Node Radius denotes the degree of the node and the Node Color the "community" it belongs to according to the Louvain method(it is essentially a grouping of the Named Entities in communitites based on the graph structure). The Links are undirected, weight-free, denoting co-occurence in same tweet between the two linked Named Entities.
-
Movies Similarity Network
The Movie Similarity Network illustrates the the similarity of movies as found from different modalities.$
-
Chord Diagram
Chord Diagram of the co-occurrences of most frequent Named Entities per Time Period. Select 1 of the 8 Time Periods from the dropdown menu. Then, the chord diagram depicts the frequency of co-occurence of the different pairs of Named Entities that were found in the data. The Named Entities are positioned on the diagram, starting from 12 o'clock and going clockwise, relative to their frequency.
-
Network of Users
This is a User-Topic Network, containing a percentage of Users as nodes. The Users were filtered based on their deegre range to make the visualization readable. The Node Radius shows the Mean Influence of User. The Node Color depends on the most proable topic of the user, based on the topic extraction methodology (15 different topics). The Links are directed links without weights.
-
Time Series
Time Series of the Mean Influence of the top-10 Named Entities for each Time Period. Select 1 of the 8 Time Periods from the dropdown menu. Then, you will see how the top-10 most frequent Named Entities, for the selected period, fared over whole time-span in terms of Mean Influence. The Mean Influence of the Named Entities is calculated as the mean of the influence of each user that that mentioned the Named Entity in their tweet.
-
Relation Extraction
The Relation Extraction Module as described. Input a text, tweet or a sentence and extract the relations contained, based on the Named Entities of the input.
-
Named Entities Network
The Named Entities Network contains a percentage of the Named Entities extracted from the tweets as nodes and linked by co-occurence. The Named Entities were selected as the most frequent ones. The Node Radius denotes the degree of the node and the Node Color the "community" it belongs to according to the Louvain method(it is essentially a grouping of the Named Entities in communitites based on the graph structure). The Links are undirected, weight-free, denoting co-occurence in same tweet between the two linked Named Entities.
-
Movies Similarity Network
The Movie Similarity Network illustrates the the similarity of movies as found from different modalities.$