Concept-Based Embeddings for Natural Language Processing

TL;DR

This paper introduces a concept-based embedding approach that combines concept-level and word-level information into a lower-dimensional space to improve various NLP tasks such as entity detection, speech recognition reranking, and sentiment analysis.

Contribution

It presents a novel embedding method that effectively integrates concept and word information for enhanced NLP task performance.

Findings

Improved named entity detection accuracy

Enhanced speech recognition reranking performance

Better sentiment analysis results

Abstract

In this work, we focus on effectively leveraging and integrating information from concept-level as well as word-level via projecting concepts and words into a lower dimensional space while retaining most critical semantics. In a broad context of opinion understanding system, we investigate the use of the fused embedding for several core NLP tasks: named entity detection and classification, automatic speech recognition reranking, and targeted sentiment analysis.