Understanding metrics and data insights

Learn about the metrics and project data provided by Affective Computing.

Affective Computing lets you gain key data insights from your projects and visualize them on dashboards. You can view data based on the following data sources:

Data source (Insight type)	Description
Real-time operations data	Live data from active interaction sessions, enabling you to receive real-time feedback (that is visually represented in data visualizations and widgets) about ongoing interactions.
Historical data	Data from past (completed) sessions visually represented in data visualizations and widgets. You can also use this data for additional analysis and integrate it with external data analytics tools such as Power BI or Tableau.
Psychological data	Distribution of data (from multiple projects) across psychological profiles. This data provides you with insights about your solution's performance (such as efficiency) with respect to the different profiles.

Insights from real-time operations data

These insights help you understand the events occurring in your solution at any given moment. You can learn about the following:

Widget or visualization	Description
Logged in users	Number of end-users of the solution who are currently logged in. The following example illustrates a widget that displays the number of drivers logged in to a transportation safety improvement solution:
Active sessions	Number of logged in end-users who are interacting correctly with the solution.
Inactive sessions	Number of logged in end-users who are not interacting correctly with the solution.
Anomaly detected	Number of users for whom EDAA™, the underlying technology that powers Affective Computing, detects a deviation from the free or imposed motivation.
Location map	Locations of the logged in users on a map. This data is available only if the location variable is selected during project setup. The following example illustrates a widget that displays the locations of logged in users:
Real-time operations status	Notifications about events and statuses in real time based on preconfigured logics. The following example illustrates a widget that displays the real-time operations status of a transportation safety improvement solution:
Unsafe patterns probability	Probability of having a result opposite to the project's set motivation. Note: This widget provides useful information in cases of imposed motivation. The following example illustrates a widget that displays the probability of undesirable patterns:
Motivated users	Percentage of efficiency that EDAA™ achieves every 15 minutes. The following example illustrates a widget that displays the percentage of motivated users for a project:

Insights from historical data

These insights help you understand and monitor the performance of your solution over time. You can learn about the following:

Widget or visualization	Description
Level of efficiency graph	Provides insights about how an action launched by EDAA™, the underlying technology that powers Affective Computing, increased or decreased anomalies during a session. It is measured as the average of all parameters that define the motivation vector. In the graph, the X axis indicates the date and the Y axis indicates the level of efficiency. The following example illustrates a widget that displays this graph for a session:
Level of engagement and micro-moment graph	Provides the following insights: Level of engagement: Indicates how well the combination of delivered actions performed in a session. Higher engagement means that the graph has a lower anomaly area. This metric enables you to detect good performance the strategy engine. The graph represents the concrete parameters or variables of the motivation versus the delivered actions. Micro-moment: Indicates the moment of high or low entropy. It combines the metrics of engagement and efficiency and pinpoints the moment when a concrete action delivered during a session generated the best result for the logic applied by the strategy engine. In the graph, the X axis indicates the date and the Y axis indicates the level of efficiency. The following example illustrates a widget that displays this graph for a session in a transportation safety improvement solution:
Heat map	Number and respective percentages of anomalies at the coordinates represented on a location map. This data is available only if the location variable is selected during project setup. The following example illustrates a widget that displays this graph in a transportation safety improvement solution:
Undesirable patterns detected	Displays the detected patterns that are opposite to the set motivation. This is applicable in cases of imposed motivation.
Actions ranking	Displays the ranking of actions to provide information about the ones that have the highest level of engagement or efficiency. The following example illustrates a widget that displays the engagement levels of different types of actions:
Attributes efficiency by profile	Provides information about which attributes have the best and worst efficiency, filtered by psychological profile type. The following example illustrates a widget that displays this graph:
Engagement checkpoints	Indicates how different physical checkpoints (from product calibration or fed as an interaction input) impact users' levels of engagement. The following example illustrates a widget that displays this graph for a neuroarchitecture solution:

Filtering historical data visualizations

The Filtering Tool allows you to customize the data visualizations for engagement, efficiency, and micro-moments. You can filter the visualizations by these variables either collectively or individually. It also lets you define the time range for viewing historical data, enhancing the specificity and relevance of the insights.

Insights from psychological data

These insights help you understand your end-users at the deepest level. You can learn about the following:

Data insight	Description
Psychological profiles	Percentage of each of the four psychological profiles for each user of the project.
Micro-moment	Indicates the best and worst moment of action delivery, filtered by psychological profile type. The following example illustrates a widget that displays micro-moments for your solution for users with self-centered psychological profiles:
Action efficiency	Indicates the best and worst action based on engagement, filtered by profile.
Action engagement	Indicates the best and worst action based on engagement, filtered by profile.
Profile anomaly	Percentage of anomalies for each psychological profile. The following example illustrates a widget that displays the percentage of anomalies for the different psychological profile types:
Personal skills	Percentage of each of the 22 symptoms on which EDAA™, the underlying technology that powers Affective Computing (powered by Virtue), bases psychological profile analysis for each user. You can select symptoms from the following that align best with your project: Good level of focus Repression Presence of stress Presence of anxiety Coherence Self-knowledge Dominant attitude High level of reactivity Presence of aggressiveness Rule-abiding Flexible mind Problem-solving Creativity Determination Ambiguity tolerance Leadership Accommodating Assertiveness Decision-making Conflict resolution Collaborative The following example illustrates a widget that displays this information for the level of focus symptom:
Anomalies (of end-users)	Percentage of anomalies in the last 15 days based on the total amount of anomalies in the project.
Profiles (of end-users)	Psychological profiles of users and their explanations.

Visualizing project data insights for your solution

You can view data insights for individual projects in the Portal, our no-code SaaS GUI platform:

Viewing project data insights

---

Related information

Data visualization