Estimating Referential Properties of Japanese Noun Phrases

Title: Estimating Referential Properties of Japanese Noun Phrases

Abstract: The research aimed to develop a machine-learning approach to estimate the referential properties of Japanese noun phrases, which are classified as generic, definite, and indefinite. This is crucial for article generation in machine translation and anaphora resolution in Japanese noun phrases. The study automated the adjustment of scores given by heuristic rules, reducing the manpower cost.

Question: How can we automate the adjustment of scores given by heuristic rules to estimate the referential properties of Japanese noun phrases?

Methodology: The study used a machine-learning method to adjust the scores given by heuristic rules. They developed a system that could learn from examples and adjust the scores accordingly. This system was trained on a dataset of Japanese sentences with annotated referential properties.

Results: The study successfully developed a machine-learning system that could automatically adjust the scores given by heuristic rules. This system reduced the manpower cost and improved the efficiency of estimating the referential properties of Japanese noun phrases.

Implications: The study's approach has significant implications for natural language processing. It provides a more efficient and accurate method for estimating the referential properties of Japanese noun phrases, which is crucial for article generation in machine translation and anaphora resolution. The study also contributes to the field by demonstrating the potential of machine learning in natural language processing tasks.

Link to Article: https://arxiv.org/abs/0103011v1 Authors: arXiv ID: 0103011v1