Automatic Recognition of Domain-Specific Terms: an Experimental Evaluation

Abstract

This paper presents an experimental evaluation of the state-of-the-art approaches for automatic term recognition based on multiple features: machine learning method and voting algorithm. We show that in most cases machine learning approach obtains the best results and needs little data for training; we also find the best subsets of all popular features.

Keywords

• автоматическое распознавание терминов
• извлечение терминов
• машинное обучение
• экспериментальная проверка
• поиск информативных признаков