Emotional Type Profiling Using Quantitative User Indicators
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for user profiling in customer management, particularly in banking, face challenges due to the qualitative nature of data collected from self-assessments, making it difficult to correlate personal data with banking dynamics, leading to high computational costs and inefficiencies in classifying large numbers of customers.
Innovation Solution
A method using quantitative indicators and optimized algorithms to classify users into emotional types, reducing computational cost by profiling based on a small user sample without acquiring personal data from each individual, utilizing partitioning and classification algorithms to identify optimal clusters representative of emotional traits.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If qualitative data from self-assessments is used for user profiling, then personal data can be collected, but it becomes difficult to correlate with user behavior and requires high computational costs
Solution Approach 1:
The patent transforms the nature of data parameters from qualitative self-assessment data to quantitative behavioral data. By changing the parameter type from subjective perceptions to objective measurable actions, the system enables efficient correlation with user behavior while reducing computational complexity through standardized numerical indicators.
Solution Approach 2:
The patent replaces the manual data collection and processing mechanism with an automated system that extracts quantitative indicators directly from user interactions. This substitution eliminates the need for complex computational processing of qualitative responses while maintaining accurate user profiling through objective behavioral metrics.
2Measurement precision
If specific personal data is acquired and processed for each user to classify them into groups, then accurate user classification is achieved, but the acquisition operation and data processing become difficult with high computational cost
Solution Approach 1:
The patent extracts only the essential quantitative indicators needed for user classification from the broader set of possible user data. By taking out only the relevant behavioral metrics rather than processing all available personal data, the system achieves accurate classification while significantly reducing data acquisition and processing complexity.
Solution Approach 2:
The patent segments the user classification process into distinct phases: first identifying quantitative behavioral indicators, then grouping users based on these indicators. This segmentation allows for simplified data collection focused on specific metrics while maintaining classification accuracy through systematic processing stages.
3Adaptability or versatility
If qualitative user data is used for profiling, then individual user characteristics can be captured, but it is poorly applicable in the management of a large number of customers
Solution Approach 1:
The patent changes the data parameters from qualitative individual characteristics to quantitative behavioral indicators that can be systematically measured and compared across large populations. This parameter transformation enables individual user characterization to be maintained while achieving scalability through standardized numerical metrics that can be processed efficiently at scale.
Data Source
Figure 1
AI summary
Method for profiling the emotional type of one or more users, comprising the steps of providing (1) a database containing a first set of data comprising for each user of a plurality of users a respective set of data, and a plurality of indicators representative of each set of data; providing (2) a processing unit in signal communication with the database; selecting (3) a sample of users; acquiring (4) a set of qualitative data for each user of the sample by means of at least one data entry form; providing (5) a partitioning algorithm; profiling (6) the users of the sample by means of the partitioning algorithm, this step comprising the sub-steps of training (61) the partitioning algorithm for each set of qualitative data, defining (62) an optimal number of clusters related to different emotional types and classifying (63) the users of the sample by assigning them to a respective cluster on the basis of the respective set of qualitative data; providing (7) a classification algorithm; profiling (8) the users of the sample of users in the clusters by means of the classification algorithm based on the indicators; evaluating (9) the classification algorithm as a function of a comparison between the profiling obtained by means of the classification algorithm and the profiling obtained by means of the partitioning algorithm, this step comprising the sub-steps of validating (91) the classification algorithm when there is compatibility in the comparison of said profilings, or repeating (92) the steps of providing (7), profiling (8) and evaluating (9) when there is no compatibility in the comparison; profiling (10) each user of the plurality of users by assigning them to a cluster, to define their emotional type, by means of the classification algorithm to which the indicators are supplied as input.