The following charts depict the detailed analysis of the FOMC20050202meeting. The full dialog of this meeting can be discovered through the following link:
Transcript for meeting FOMC20050202meeting
.
The first two charts visualize the speaker adjacency and speaker similarity for the whole
meeting and all participating speakers. The following diagrams depict the analysis of each subdialog in the meeting.
To learn more about the underlying concepts for this
analysis please refer to the full
publication
.
The following chord diagram visualizes the speaker relations for all speakers participating the meeting and for the whole meeting. A relation between speaker A and B means that speaker B is talking after speaker A or vice versa. We connect speaker A and B with an edge, while the weight of an edge is the number of those relations occurring throughout the meeting. The visualization contains interactive elements that allow to highlight specific speakers or relations and display additional information (For example X% of Speaker A after Speaker B" means that X % of all turns of speaker A are subsequent to speaker B).
The following heatmap visualization depicts the average speaker similarity values for this meeting, displaying the cosine similarities for each speaker pair through a dynamic color scale. The average similarities of selected speakers over all meetings can be found on the overview page.
The following chord diagrams visualize the speaker relations for all speakers of a meeting subdialog. To learn more about the
concept of subdialog identification please refer to chapter 5 of the full
publication (not yet available)
. The transcript of each subdialog can be accessed through the link below the dialog title.
A relation between speaker A and B means that
speaker B is talking after speaker A or vice versa. We connect
speaker A and B with an edge, while the weight
of an edge is the number of those relations occurring throughout the meeting.
The visualization contains interactive elements that allow to highlight specific speakers or relations and display additional information (For example
X% of Speaker A after Speaker B" means that X % of all turns of speaker A are subsequent to speaker B).