Contextual Relationship Graphs for Targeted Content Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for providing targeted content, such as targeted advertising on web pages, face challenges in accurately classifying user context due to their reliance on context-independent analyses or slower, human-intensive natural language processing, which limits their ability to provide strong contextual relationships and efficient real-time analysis.
Innovation Solution
A system and method utilizing a hierarchical predictive projection with contextual relationship graphs to select targeted content based on request-associated attributes, enabling efficient and accurate contextual analysis without human intervention, by analyzing relationships between keywords and determining the relative strength of classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the bucket of words approach is used for context classification, then the processing speed is fast and it can be applied in real-time, but the accuracy of contextual analysis is poor because words are analyzed without regard to context and relationship to other words
Solution Approach 1:
The patent introduces contextual relationship graphs as an intermediary structure that captures semantic relationships between words. These graphs serve as a mediator between the simple bucket of words approach and the need for contextual understanding, enabling fast processing while improving accuracy through pre-computed contextual relationships.
Solution Approach 2:
The patent pre-computes and stores contextual relationships between words in graphical structures before the actual content classification task. This preliminary action allows the system to quickly query pre-established contextual relationships during real-time processing without performing complex contextual analysis on the fly.
2Measurement precision
If natural language processing is used for context classification, then the accuracy of contextual analysis is improved through contextual extraction, but the processing speed decreases significantly and it cannot be used for online real-time analysis
Solution Approach 1:
The patent extracts only the essential contextual relationships between words and stores them in graphical structures, rather than performing complete natural language processing. This extraction approach captures the most important contextual information while avoiding the computational overhead of full NLP analysis.
Solution Approach 2:
The patent creates simplified graphical representations (contextual relationship graphs) that copy and store the essential contextual patterns from natural language. These graphs serve as lightweight copies that can be quickly queried without requiring the full computational resources of NLP processing.
3Measurement precision
If human involvement is added to deal with non-congruous classification trees, then the accuracy of matching web pages and advertisements is improved, but the efficiency of the process decreases further
Solution Approach 1:
The patent enables the system to automatically handle classification tree matching by using contextual relationship graphs to identify and resolve non-congruities. The system serves itself by computationally detecting and correcting classification mismatches without requiring human intervention, thereby maintaining high accuracy while preserving automation and efficiency.
4Device complexity
If context-independent analysis is used, then the simplicity of implementation is maintained, but the strength of contextual relationships is inadequate and results are skewed by inadequate or false information
Solution Approach 1:
The patent nests contextual relationship graphs within the existing content classification system. The graphical structures are embedded as an additional layer of contextual information that enhances the simple bucket of words approach without fundamentally redesigning the entire system, thus maintaining implementation simplicity while improving reliability.
Data Source
AI summary
An arrangement for providing targeted content includes data repositories storing information from which targeted content may be selected. The data repositories store at least one contextual relationship graph. The arrangement also includes an input/output interface through which a request for targeted content is made. The arrangement further includes a controller that receives the request for targeted content and selects targeted content using the contextual relationship graph. The controller further provides the selected targeted content through the input/output interface. An arrangement for determining the relative strength of a classification for a group of words includes memory for storing a contextual relationship graph for a given classification and a processor that receives the contextual relationship graph and a plurality of words to be analyzed by the processor, identifies occurrences of the relationships identified in the contextual relationship graph and determines the relative strength of classification based on the identified occurrences.