Organization Classification via Transactional Data Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for classifying organizations into product-oriented, service-oriented, and project-oriented categories are inadequate, as they rely on self-reported data, sales channels, or MCC codes, which can be unreliable, especially for organizations that do not provide clear type information or use unconventional sales channels.
Innovation Solution
A method and system that classify organizations based on transactional data using supervised learning, where descriptive texts from transactions are processed to generate vectors, and a classifier is trained to predict the organization type, allowing for accurate categorization without relying on self-reported data or conventional sales channels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If self-reported data or MCC codes are used for organization classification, then the classification process is simple, but the reliability of classification is low
Solution Approach 1:
The patent replaces traditional mechanical classification methods (self-reported data, MCC codes) with a machine learning-based system that automatically analyzes transactional data. The classifier uses supervised learning to process descriptive texts from transactions, converting unstructured data into structured vectors for accurate organization type classification, thereby substituting manual or rule-based approaches with an automated intelligent system.
Solution Approach 2:
The patent introduces an intermediary processing layer that transforms raw transactional data into meaningful features. The system uses vectorization of descriptive texts and frequency-based feature extraction as intermediaries between the raw data and the final classification decision, enabling the classifier to indirectly assess organization type through patterns in transaction descriptions rather than direct self-reporting.
2Measurement precision
If transactional data processing is performed to improve classification accuracy, then measurement precision improves, but loss of time increases
Solution Approach 1:
The patent extracts only the most relevant features from transactional data for classification. Instead of processing entire transactional datasets, the system identifies and extracts key features such as frequent words in descriptive texts and converts them into vectors. This selective extraction maintains high classification precision while significantly reducing the time and computational resources required compared to analyzing complete transaction histories.
Solution Approach 2:
The patent applies partial action by using a subset of transactional data features rather than the complete dataset. The system processes only the descriptive text portions of transactions and extracts frequency-based features, performing a partial analysis that achieves sufficient classification accuracy without the excessive time cost of comprehensive transactional data processing.
Data Source
AI summary
A method for classifying organizations involves obtaining, for an unknown organization, transactional data representing a multitude of transactions. The transactional data comprises a descriptive text for each of the multitude of transactions. The method further involves processing the descriptive text for each of the multitude of transactions to obtain one vector representing the unknown organization, categorizing the unknown organization using a classifier applied to the vector, and identifying a software service for the unknown organization, according to the categorization.