Related Concept Videos
Components of Language
Causes of Similarity-Dissimilarity Effect
Natural and Artificial Concepts
Automatic Processing and Automatic Social Behavior
Genome Annotation and Assembly
Improving Translational Accuracy
You might also read
Related Articles
Articles linked to this work by shared authors, journal, and citation graph.
Research on Chinese Speech Emotion Recognition Based on Deep Neural Network and Acoustic Features.
Applying attention-based BiLSTM and technical indicators in the design and performance analysis of stock trading strategies.
Related Experiment Video
Updated: Apr 27, 2026

Experimental Paradigm for Measuring the Effect of Induced Emotion on Grammar Learning
Published on: January 29, 2020
A grammar-based semantic similarity algorithm for natural language sentences.
Ming Che Lee1, Jia Wei Chang2, Tung Cheng Hsieh3
1Department of Computer and Communication Engineering, Ming Chuan University, Taoyuan 333, Taiwan.
This study introduces a novel sentence similarity algorithm using corpus-based ontology and grammar rules. It significantly improves natural language understanding for texts lacking obvious conceptual overlap.
Area of Science:
- Natural Language Processing
- Computational Linguistics
- Information Retrieval
Background:
- Traditional information retrieval methods struggle with sentence similarity when conceptual overlap is minimal.
- Existing approaches like vector models and ontology-based methods have limitations in accurately matching sentences with arbitrary structures.
Purpose of the Study:
- To propose a novel sentence similarity algorithm for natural language.
- To address the limitations of traditional methods in identifying semantic similarity between sentences lacking explicit concept overlap.
Main Methods:
- Developed a similarity algorithm leveraging corpus-based ontology and grammatical rules.
- Utilized a grammar and semantic corpus for training and evaluation.
Main Results:
- The proposed algorithm demonstrates significant performance improvements on benchmark datasets.
- Achieved enhanced accuracy in matching sentences and short texts with diverse syntax and structure.
Conclusions:
- The grammar and semantic corpus-based algorithm offers a robust solution for natural language sentence similarity.
- This approach effectively overcomes limitations of traditional methods, particularly for texts with no obvious semantic relation.