Conversational Query Response Using Structured-to-Text Data Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Dialogue systems face challenges in processing mixed structured and unstructured data, leading to difficulties in identifying information and requiring extensive training resources and time for neural networks.
Innovation Solution
A system that generates contextual data using both structured and unstructured data formats, allowing neural networks to process input requests without additional training, by converting structured data into unstructured form and combining it with unstructured data to generate responses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If structured data is used in knowledge databases, then information can be organized with specific identifiers, but training neural networks requires large amounts of training data for each identifier, increasing computing resources and time
Solution Approach 1:
The patent changes the format parameter of structured data by converting it into unstructured text format. This transformation allows the dialogue system to process information without requiring extensive training on each structured identifier, thereby reducing training time while maintaining information organization through the converted text representations.
Solution Approach 2:
The patent creates text-based copies of structured data fields and converts them into unstructured format. These copied and transformed data representations enable the neural network to learn from natural language patterns rather than memorizing structured identifier mappings, significantly reducing the training data requirements and computing resources needed.
2Quantity of substance
If both structured data and unstructured data are included in knowledge databases, then comprehensive information is available, but it becomes difficult for the dialogue manager to identify needed information due to different representations
Solution Approach 1:
The patent merges structured data and unstructured data by converting the structured data into unstructured text format. This unification creates a consistent data representation that eliminates the difficulty of identifying information across different formats, while preserving the completeness of information from both structured and unstructured sources.
3Productivity
If structured data is converted to unstructured form and combined with unstructured data, then less training data and computing resources are needed, but the system must process mixed data formats
Solution Approach 1:
The patent extracts structured data from its original format and separates it for conversion into unstructured text. This extraction process allows the system to transform only the necessary structured portions while maintaining the existing unstructured data, thereby improving training efficiency without significantly increasing overall data processing complexity.
Data Source
AI summary
In various examples, contextual data may be generated using structured and unstructured data for conversational AI systems and applications. Systems and methods are disclosed that use structured data (converted to unstructured form) and unstructured data, such as from a knowledge database(s), to generate contextual data. For instance, the contextual data may represent text (e.g., narratives), where a first portion of the text is generated using the structured data and a second portion of the text is generated using the unstructured data. The systems and methods may then use a neural network(s), such as a neural network(s) associated with a dialogue manager, to process input data representing a request (e.g., a query) and the contextual data in order to generate a response to the request. For instance, if the request includes a query for information associated with a topic, the neural network(s) may generate a response that includes the requested information.