In this article:
Using Heartbeat involves first defining the themes that you would like to track. Themes are then applied to time-series text data to create a Heartbeat chart that visualizes how those discussions are changing over time.
The best approach to creating themes will depend on whether you already know what you want to track.
What are themes?
Themes are collections of topics, words, phrases and grammar elements that you want to track on an ongoing basis. Within a theme, you can track either individual linguistic features or group related terms together into sub-themes.
Themes are applied to time-stamped data to create Heartbeat charts that visualize how the proportion of conversation dedicated to your themes is changing over time.
You define a theme group titled 'Electric vehicles.' Within this theme group, there may be several themes relating to different elements of the conversation. For example:
Appearance - attractive, beautiful, sporty, fancy, cool
Affordability - cheap, affordable, expensive, cost
Supply chain - lithium, mining
Each of these themes would display as its own line on a Heartbeat chart.
When you need to identify what to track
For the most effective results and robust insights, we recommend creating a comparison in Explore to identify the key features of your data set to track on an ongoing basis.
When you find something interesting in the results of your comparison, launch the theme builder side panel to create a new theme group or add to an existing one
Create and name a new theme group or select the flag icon to add to an existing theme group
Drag and drop the features you want to include as part of the theme into the side panel.
Click 'Create Heartbeat chart'
When you already know what you want to track
Select 'Heartbeat' from the platform menu then click on the 'Themes' tab
Click the blue plus button in the top right and name your theme
Hover over the theme and select 'Add item', then use the drop-down to select the attribute type you want to add and type in the words and phrases or select the topics or grammar elements that you want to monitor
Drag and drop to group related linguistic features into sub-themes if desired.
Now you're ready to apply your themes to your data to create a Heartbeat chart.