Context-Aware Text Representation for Social Relation Aided Sentiment Analysis

被引：2

作者：

Nguyen, Minh Luan ^{[1
]}

机构：

[1] Inst Infocomm Res, Singapore, Singapore

来源：

PROCEEDINGS OF THE 25TH INTERNATIONAL CONFERENCE ON WORLD WIDE WEB (WWW'16 COMPANION) | 2016年

关键词：

sentiment analysis; deep learning; text embedding; social relations;

D O I：

10.1145/2872518/2889347

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

In this paper, we propose CaTER, which learns a novel context-aware joint representation of text and user by incorporating semantic text embedding of unlabeled tweets as well as social relation information. CaTER leverages the wealth of user contextual information available apart from user's utterances for sentiment analysis. Our approach is inspired by social science about emotional behaviors of connected users, who perhaps more likely to consensus on similar opinions. Our method outperforms numerous baselines on two real-world Twitter datasets.

引用

页码：85 / 86

页数：2