Semantic Target Prediction Using Structured and Unstructured Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing future-casting technologies face limitations in accurately predicting targets due to the challenge of integrating structured and unstructured data, particularly in predicting numerical values, and lack of clear basis for causal relationships between targets and influence variables.
Innovation Solution
A target prediction method and system that utilizes a language model to detect target influence variables at a semantic level, filters structured and unstructured data, and generates relationship information to predict short-term to long-term outlooks by integrating data through a time series analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If machine learning models are used to predict future events using past data, then prediction capability is improved, but accuracy is limited by data quality and quantity
Solution Approach 1:
The patent segments the prediction system into multiple specialized components: a language model processor for unstructured text data, a structured data processor for numerical data, and a integration module. This segmentation allows each component to optimize for its specific data type, improving overall prediction accuracy while maintaining robust prediction capability across diverse data sources.
Solution Approach 2:
The patent creates a composite prediction system that integrates multiple data types (structured and unstructured) and multiple processing approaches (statistical analysis and language model-based analysis). This composite approach combines the strengths of different methods to overcome the limitations of any single approach, thereby improving prediction accuracy without sacrificing reliability.
2Productivity
If only structured data is used for prediction, then data processing efficiency is improved, but ability to utilize unstructured data is lost
Solution Approach 1:
The patent introduces a language model as an intermediary that translates unstructured text data into structured numerical representations. This intermediary enables the efficient processing pipeline to handle both structured and unstructured data uniformly, maintaining processing efficiency while dramatically expanding data type versatility.
Solution Approach 2:
The patent transforms unstructured text data into structured numerical parameters through language model processing. By changing the parameter representation from unstructured text to structured numerical values, the system maintains efficient processing speeds while gaining the ability to utilize previously unusable unstructured data sources.
3Adaptability or versatility
If future-casting technologies are applied to predict numerical values, then prediction scope is expanded, but accuracy deteriorates due to integration challenges
Solution Approach 1:
The patent segments the data integration process into distinct stages: language model processing for unstructured data, structured data processing for numerical data, and a dedicated integration module. This segmentation allows each stage to optimize for its specific data type while maintaining overall numerical prediction accuracy, thereby expanding prediction scope without sacrificing precision.
4Measurement precision
If semantic level analysis is performed to detect target influence variables, then prediction basis is improved, but computational complexity increases
Solution Approach 1:
The patent performs preliminary semantic analysis to identify and extract target influence variables before the main prediction process. By pre-processing the data to isolate key influential variables, the system improves prediction basis quality while reducing the computational complexity of the subsequent prediction steps, as the model only needs to process the extracted key variables rather than all raw data.
Data Source
AI summary
A target prediction method for predicting a future outlook of a target performed by a computing device or a processor may collect related structured and unstructured data when a user requests target prediction, analyze the relationship between the target and a variable affecting the target at a semantic level, and compute a target outlook of a future.