Business Classification Engine Using Power Consumption Profiles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods lack an efficient way to classify businesses based on physical and behavioral attributes for improved utility services and demand response notification systems, which affects the relevance of advertising and notification content.
Innovation Solution
A computer-implemented method and system that receives business categories with associated category profiles, compares business information to these profiles using physical and behavioral attributes, and assigns classifications to unclassified businesses, employing clustering algorithms and machine-learning approaches to determine similarity and update classifications as needed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional classification methods are used for businesses, then the system is simple to operate, but the classification precision and relevance of utility services are insufficient
Solution Approach 1:
The patent segments business classification into multiple dimensions including physical attributes (location, size, age) and behavioral attributes (power consumption patterns, operational hours). This segmentation allows for more precise classification by evaluating businesses across multiple independent criteria rather than using a single monolithic classification method.
Solution Approach 2:
The patent implements dynamic classification by continuously monitoring behavioral attributes such as power consumption patterns and updating business classifications in real-time. The system adapts to changing business characteristics and adjusts classifications dynamically, rather than using static categorization methods.
2Loss of information
If detailed physical and behavioral attributes are collected for businesses, then the relevance of demand response notifications is improved, but the amount of data processing and system complexity increases
Solution Approach 1:
The patent extracts specific relevant attributes from comprehensive business data, focusing on key physical attributes (location, size, age) and behavioral attributes (power consumption patterns, operational hours). By selecting and extracting only the most relevant features rather than processing all available data, the system maintains high information relevance while reducing processing complexity.
Solution Approach 2:
The patent introduces category profiles as intermediary structures that mediate between raw business attributes and classification decisions. These profiles serve as templates containing expected attribute patterns for different business types, simplifying the matching process by providing pre-defined classification criteria rather than requiring complex real-time analysis of all attributes.
3Reliability
If machine-learning algorithms are used to update business classifications, then the classification accuracy is improved, but the computational resources and processing time are increased
Solution Approach 1:
The patent implements partial machine-learning application by using clustering algorithms only for initial business categorization and pattern recognition, rather than applying complex machine-learning models to all classification decisions. This selective application of machine-learning provides sufficient accuracy for the use case while maintaining processing efficiency.
Solution Approach 2:
The patent enables self-service classification by allowing the system to automatically update business classifications based on monitored behavioral attributes without requiring manual intervention or complex retraining of machine-learning models. The system autonomously adjusts classifications as businesses change their power consumption patterns or operational characteristics.
Data Source
AI summary
Aspects of the subject technology relate to methods and systems for classifying businesses based on various types of information, such as resource consumption information. In some implementations, methods of the subject technology include steps for receiving a plurality of business categories, wherein each of the business categories is associated with at least one category profile, and receiving business information for an unclassified business, wherein the business information comprises behavioral attribute information corresponding with the unclassified business. In some implementations, the methods disclosed herein can further include steps for comparing the business information to one or more of the category profiles to determine if the unclassified business should be associated with at least one of the plurality of business categories.