NLG System Automatic Referential Expression Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional natural language generation (NLG) systems require users to specify complex rules for using referential and anaphoric expressions, leading to time-consuming processes and often generate text with ambiguous references.
Innovation Solution
An NLG system that automatically determines the appropriate language for referring to referents by accessing information on referential and anaphoric expressions and using system rules to choose between them, avoiding ambiguous references.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If users specify complex rules for using referential and anaphoric expressions in conventional NLG systems, then the system can generate text with controlled reference usage, but the process becomes time-consuming and requires significant user effort
Solution Approach 1:
The system automatically determines whether to use referential or anaphoric expressions by analyzing the text context and referent information, eliminating the need for users to manually specify complex rules. The NLG system serves itself by making autonomous decisions about reference expression selection based on predefined criteria and contextual analysis.
Solution Approach 2:
The system pre-processes information about referents and potential reference expressions before text generation, organizing data about referential and anaphoric expressions in advance. This preliminary organization allows the system to quickly select appropriate expressions during text generation without requiring users to specify detailed rules for each case.
2Ease of operation
If conventional NLG systems use simple template-based techniques, then the text generation process is straightforward, but the generated text may contain ambiguous references
Solution Approach 1:
The system incorporates feedback mechanisms that analyze the generated text context to determine whether referential or anaphoric expressions will produce ambiguous references. By continuously monitoring contextual factors and adjusting expression selection based on this feedback, the system maintains clarity while using straightforward template-based generation.
Solution Approach 2:
The system dynamically changes parameters such as the type of reference expression (referential vs. anaphoric) based on contextual analysis. By adjusting these linguistic parameters according to the specific context, the system generates clear references while maintaining the simplicity of template-based text generation.
3Productivity
If the system automatically selects between referential and anaphoric expressions using system rules, then user burden is reduced and text generation is faster, but the system must process and analyze more information to make accurate selections
Solution Approach 1:
The system segments the information processing task into distinct components: identifying referents, analyzing contextual factors, selecting appropriate reference expressions, and generating text. By dividing the complex processing into manageable segments, the system reduces the cognitive load at each stage while maintaining overall accuracy and speed.
Solution Approach 2:
The system uses intermediary structures such as predefined criteria, contextual analysis frameworks, and selection algorithms that mediate between raw information and final expression selection. These intermediaries organize and structure the information processing, making it more efficient and manageable while enabling automatic decision-making.
Data Source
AI summary
Techniques for use in connection with a system for automatically generating text. Techniques include accessing information specifying at least one referential expression for at least a first referent and at least one anaphoric expression for at least the first referent; accessing a template that includes human-language text and a first tag that serves as a placeholder for a first text portion including a reference to at least the first referent; automatically identifying, using at least one system rule and at least one processor, text to use for the first text portion at least in part by determining whether to use as the text for the first text portion the at least one referential expression or the at least one anaphoric expression; and automatically generating output text including the human-language text and the identified text for the first text portion.