Dynamic Phone Number Whitelists From Call Data for Spam Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to accurately and efficiently manage phone number blacklists and whitelists to reduce spam calls, relying on time-consuming data analysis and varying accuracy across different called parties, which affects customer satisfaction.
Innovation Solution
A server-based system that analyzes call data, user activity, and device locations to generate and manage network whitelists and blacklists, synchronizing them with local lists and merging them as needed, using machine learning to improve accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing systems use data analysis to manage phone number blacklists and whitelists, then they can reduce spam calls, but the process is time-consuming and accuracy varies across different called parties
Solution Approach 1:
The system pre-processes and stores call data, user activity patterns, and device location information in databases before actual call assessment is needed. This preliminary data collection and organization enables rapid retrieval and analysis during real-time call decision-making, reducing the time required for accurate spam call determination.
Solution Approach 2:
The patent replaces manual or simple filtering mechanisms with automated machine learning algorithms that process multiple data dimensions simultaneously. The system uses computational models to analyze call patterns, user behavior, and contextual information, substituting time-consuming manual analysis with efficient automated processing that maintains high accuracy.
2Measurement precision
If the system analyzes multiple factors for each phone number to generate accurate whitelists and blacklists, then accuracy improves, but the complexity of data collection and processing increases
Solution Approach 1:
The system divides the complex data analysis task into separate modular components: data collection modules for call data, user activity, and device location; data processing modules that analyze each factor independently; and decision modules that integrate results. This segmentation allows accurate multi-factor analysis while reducing overall system complexity through modular architecture.
Solution Approach 2:
The patent introduces intermediary data structures and processing layers that bridge raw data collection and final classification decisions. These intermediaries organize and pre-process multiple factors into structured formats that can be efficiently analyzed, reducing the complexity of direct multi-factor processing while maintaining measurement precision.
3Reliability
If the system uses machine learning to improve accuracy, then spam call reduction effectiveness increases, but the computational resources and processing time required increase
Solution Approach 1:
The system performs machine learning model training and feature importance analysis in advance, pre-processing data to identify and prioritize the most significant factors for spam detection. This preliminary work enables the system to use computational resources more efficiently during actual call assessment by focusing only on the most impactful features rather than analyzing all data equally.
Solution Approach 2:
The patent applies machine learning selectively rather than universally - using complex models only for high-risk or ambiguous call scenarios while employing simpler rules for clear-cut cases. This partial application of machine learning maintains high effectiveness for difficult spam detection while reducing overall computational resource consumption compared to applying complex algorithms to all calls.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for managing a phone number list are disclosed. In one aspect, a method includes the actions of receiving, by a computing device, telephone call data that reflects characteristics of telephone calls placed and received by a first user. Based on the telephone call data, the actions further include generating, by the computing device, a first telephone number whitelist for the first user. The actions further include determining, by the computing device, whether to combine the first telephone number whitelist and a second telephone number whitelist for a second user. The actions further include storing, by the computing device, the first telephone number whitelist or the combined telephone number whitelist in association with the first user.