Content Selection System Using Sensor Network Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face difficulty in identifying relevant content from the vast array of accessible content due to conventional methods that only capture limited aspects of user behavior, such as internet browsing habits, failing to tailor content effectively to individual interests and needs.
Innovation Solution
A system and method that utilize sensor network data to compare item usage at a location with content attributes, selecting and transmitting a subset of content to client devices based on usage patterns and interaction history while maintaining privacy settings, ensuring relevance and personalization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If conventional methods (internet browsing habits) are used to identify relevant content, then the system complexity is low, but the content relevance to user interests and needs deteriorates
Solution Approach 1:
The patent segments user behavior data into multiple dimensions: sensor network data (location, activity), social network data (connections, interactions), and browsing history. This segmentation allows comprehensive content personalization without requiring a monolithic complex system, as each data type can be processed independently and combined for content recommendation.
Solution Approach 2:
The patent introduces a content identification system as an intermediary that aggregates and analyzes multiple data sources (sensors, social networks, browsing history) to bridge the gap between user behavior and content selection. This intermediary processes diverse inputs and transforms them into personalized content recommendations, improving relevance without exposing users to raw data complexity.
2Loss of information
If sensor network data is collected to improve content personalization, then content relevance improves, but user privacy deteriorates
Solution Approach 1:
The patent extracts and utilizes only the necessary minimal set of sensor data elements required for content personalization (such as location and activity type) while excluding sensitive personal information. This extraction approach maintains content relevance by using sufficient behavioral data without compromising user privacy through excessive data collection.
Solution Approach 2:
The patent transforms raw sensor data into aggregated behavioral patterns and preferences rather than storing or processing individual sensitive data points. By changing the parameter representation from specific personal information to generalized usage patterns, the system maintains analytical value for content personalization while reducing privacy risks associated with storing detailed personal data.
Data Source
AI summary
Determination of content for presentation by a client device based on item usage data captured by a sensor network is disclosed. Data describing usage of one or more items at a location is received from a sensor network associated with the location. Content is received from a server and a subset of the received content is selected based on attributes of the data from the sensor network and attributes of the content. The subset of the received content is transmitted to a client device for presentation. In one embodiment, data describing interaction with the subset of the received content is received from the client device and transmitted to a content distribution server for use in selecting additional content. In an embodiment, second content determined by the server using interaction with the subset of the received content and data from the sensor network is received from the server.