Depth profiling

Understand how Affective Computing (powered by Virtue) determines the psychological profiles of end-users.

Depth profiling has two aspects:

Advantage

Depth profiling offers you nuanced insights into end-user psychology, facilitating personalized interactions and strategic decision-making (based on diverse psychological profiles) in solutions.

Affective Computing (powered by Virtue) can deliver unmatched personalization of human-machine interactions.

Although the market is full of tools that personalize interactions, the interactions they support happen at a very shallow level. For example, they base their personalization on video viewing history or search history.

Therefore, these solutions do not achieve true personalization because they are exposed to only a very shallow behavioral pattern, namely historical activity, of their users.

Using historical data to define a human user is equivalent to using a footprint on sand to define a human who walked somewhere.

Unlike these solutions, Affective Computing (powered by Virtue), with the help of its underlying technology, EDAA™, knows the end-user at the psychological level. When an end-user interacts with any solution powered by Affective Computing, EDAA™ receives and processes three types of data:

Physiological
Behavioral (user motion)
Environmental

The solution's interactions with users are personalized taking all three types of input into consideration.

Affective Computing (powered by Virtue) uses physiological inputs to create the psychological profile of the end-user (through the diagnostics process).

All solutions powered by Affective Computing involve depth profiling of the end-users. This is the core of our offering as it brings emotions into human-centric solution design.

Use case

The approach of grouping users by psychological profile can provide valuable insights for shaping company and business strategies that consider the diverse psychological makeup of employees or clients, in addition to aiding in personalizing user interactions.

How it works

Depth profiling works in parallel with the Metacognition engine (the Sensations engine and Strategy engine) to predict the user’s profile.

Depth profiling operates with the following concepts:

Symptoms: Specific signs extracted from each user's behavior during the entire or specific subsets of the solution experience that reveal their profiles. The analysis is based on our science-based methodology of calculating a person's psychological profile; the symptoms are based on mathematical variables and work as a standard reference.
Pathologies: Cognitive patterns that manifest in certain behavior that define the person’s antipathies.
Personality states: Different aspects of personality, such as awareness, identity, and reason or motivation for success and failure.

Psychological profile generation

The psychological profile similar to the DNA of the user's profile.

Symptoms (variables) are building blocks of this DNA. EDAA™ repeatedly analyzes the different data inputs available for each user and determines the percentage of each symptom in the total psychological profile DNA. It also determines whether any new symptom needs to be added at any given time based on the user's current psychological state.

Additionally, its accuracy in detecting and predicting these symptoms is also continuously increased by its Artificial Neural Network (ANN), which continues to try to find new combinations and hypotheses.

The resulting options form the learning process, which serves to improve the symptoms' standards. After the symptoms are checked and validated, they can be included in the analysis.

Symptoms provide information about the type of pathology, causal locus, and identity status of the user.

The combination of different patterns makes up the depth profiling of the user.

Parts of the psychological profile are connected through applying reinforcement learning, considering things like the profile itself, the session, the situation, and more.

The process of diagnostics establishes the initial (baseline or preliminary) psychological profile of the user. In subsequent interactions with the same user, through the process of user calibration, EDAA™ updates the profile.

For detailed information about psychological analysis and the different types of user profiles, see How EDAA™ analyzes psychological profiles. For a deeper understanding of diagnostics and calibration, see Diagnostics and calibration.

The profile's accuracy is assessed through the percentage of a metric called Engagement, which is directly linked to the process of diagnostics. If the engagement of an action delivered as a result of a logic in an experiential scenario is lower than 60%, the system is “punished” through the Feedback functionality. Otherwise, the system is rewarded based on a scale where 60% is the lowest and 100% is the highest.

Persona profiling

Persona profiling, though not directly a part of diagnostics, is a functionality that allows Affective Computing (powered by Virtue) to gain additional understanding about types of users through diagnostic questions.

For example, a solution that automates personalization of human interactions with cars might target two types of users: Car Enthusiast and Racer. While these user types do not play a role in determining the users' psychological profiles, these are parts of their personas. As actions can be attributed according to persona types, this functionality plays a role in personalization.

Responses to persona-related questions are also considered in analytics and data visualization.

How EDAA™ analyzes psychological profiles

With the help of EDAA™, its underlying technology, Affective Computing (powered by Virtue) determines psychological profiles of users through its internal processes of diagnostics and calibration.

For its analysis, EDAA™ uses 3 types of data:

The user’s physiological data
The user’s motion or behavior
The environmental context during the interaction (for example, location, time, or weather)

How EDAA™ applies our scientific methodology in its analysis

When analyzing each user and establishing and updating their psychological profile, EDAA™ does the following:

First, EDAA™ determines the degree of presence of 22 symptoms (for example, level of stress) and assigns a percentage value to each as its level. To understand more about symptoms, see Psychological profile generation.
Then, based on the established levels of the symptoms. EDAA™ determines the degree of predominance of 8 pathologies (for example, inertia). It assigns a percentage value to each pathology, where a higher value indicates a higher presence of the pathology.
Then, based on the user's pathologies, EDAA™ determines the levels of presence of 4 identity types and identifies the user's causal locus.
Finally, based on the user's identity and locus, EDAA™ identifies the user's personality type and classifies the infinite combinations into 4 categories (to simplify your understanding as a client). These 4 categories are based on predominant personality aspects.

Psychological profile categories

Based on data values from over 200 psychological variables, EDAA™ can thus generate infinite combinations of user profiles.

As our clients often prefer not to dive deep into the psychological aspects that drive our method, for simplicity, EDAA™ classifies the infinite combinations into 4 categories, which are synonymous with predominant personality aspects.

Recent academic research that studied responses from 1.5 million subjects worldwide) has revealed that each person has one of four personality types. The four categories into which EDAA™ classifies profiles were selected based on this study.

These categories are: