Ontology-Driven Dictionary Generation for NLP Ambiguity Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing natural language processing methods rely on simplistic word dictionaries and extensive grammatical rules, leading to inefficient ambiguity resolution due to the need to apply hundreds or thousands of rules to each word, resulting in prolonged processing times.
Innovation Solution
The use of an ontological hierarchy of classes and sub-classes, along with grammatical rules stored in computer memory, allows for annotation of words based on their possible classes and sub-classes, eliminating ambiguities by applying only relevant rules associated with the ontology, reducing processing time and enabling efficient ambiguity resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If hundreds or thousands of grammatical rules are applied to each word for annotation, then ambiguity resolution accuracy is improved, but processing time increases significantly
Solution Approach 1:
The patent segments the grammatical rules into two categories: structure-independent rules (applied to all words) and structure-dependent rules (applied only when syntactic structure is confirmed). This segmentation allows the system to apply fewer rules initially, reducing processing time while maintaining accuracy through selective application of additional rules when needed.
Solution Approach 2:
The patent performs preliminary annotation using structure-independent grammatical rules before syntactic parsing. This preliminary action assigns basic grammatical tags to all words immediately, providing a baseline annotation that reduces the need for extensive rule application later, thereby reducing overall processing time while maintaining accuracy.
2Adaptability or versatility
If a comprehensive dictionary with all possible word classes is used for annotation, then annotation completeness is improved, but device complexity increases
Solution Approach 1:
The patent segments the dictionary into structure-independent entries (containing basic grammatical class information applicable to all contexts) and structure-dependent entries (containing context-specific grammatical information). This segmentation allows the system to maintain a comprehensive dictionary while organizing it in a way that reduces processing complexity by only accessing relevant portions based on syntactic structure.
Solution Approach 2:
The patent applies different levels of dictionary detail to different words based on their syntactic context. Words in ambiguous contexts receive more detailed dictionary consultation, while words in clear contexts use simplified lookup. This local quality approach maintains annotation completeness where needed while reducing overall system complexity.
3Measurement precision
If extensive post-processing with grammatical rules is performed to eliminate redundant annotations, then annotation accuracy is improved, but productivity decreases
Solution Approach 1:
The patent performs preliminary grammatical rule application during the initial annotation phase, before syntactic parsing. This preliminary action establishes a foundation of annotations that reduces the need for extensive post-processing, thereby improving productivity while maintaining accuracy through subsequent selective refinement.
Solution Approach 2:
The patent implements a dynamic rule application strategy where the extent of post-processing depends on the ambiguity level detected during parsing. High-ambiguity cases receive extensive post-processing, while low-ambiguity cases receive minimal processing. This dynamic approach maintains accuracy for critical cases while improving overall productivity.
Data Source
AI summary
Natural language processing ambiguity resolution includes storing an ontological library specifying a set of grammatical rules associated with ontological classes and sub-classes. A dictionary for each class and sub-class is generated by compiling all word instances belonging to the class or sub-class. A phrase is received having at least one ambiguous word that is associated with at least one class and at least two sub-classes. The ambiguous word is annotated with the ontological classes and sub-classes to which the ambiguous word belongs and the grammatical rules associated with the ontological classes and sub-classes to which the ambiguous word belongs are retrieved from the ontological library. One or more ontological classes or sub-classes are eliminated based on the retrieved grammatical rules and a surviving ontological class and sub-class are determined to be an accurate ontological class and sub-class for the ambiguous word to resolve the natural language processing ambiguity.