人工智能培训

搜索

机器学习论文:DEPECHEMOOD ++:通过简单而强大的技术构建的双语情感词典(DepecheMood++: a Bilingual Emotion Lex

[复制链接]
tetra 发表于 2018-10-11 09:01:02 | 显示全部楼层 |阅读模式
tetra 2018-10-11 09:01:02 194 0 显示全部楼层
机器学习论文:DEPECHEMOOD ++:通过简单而强大的技术构建的双语情感词典(DepecheMood++: a Bilingual Emotion Lexicon Built Through Simple Yet  Powerful Techniques)已经开发了几种用于情感分析的词汇,并在NLP社区中提供。虽然其中大多数都带有单词极性注释(例如正面/负面),但是为了更细粒度的情感分析(例如幸福,悲伤)而建立lexica的尝试最近引起了极大的关注。这种词汇经常被用作开发学习模型的过程中的构建块,其中需要情感识别,和/或用作比较模型性能的基线。在这项工作中,我们为社区贡献了两个新的资源:a)扩展了现有和广泛使用的英语情感词典;和b)针对意大利语的新词典。此外,我们展示了在监督和非监督实验环境中如何使用简单技术来提高数据集和不同程度域特异性任务的性能。
Several lexica for sentiment analysis have been developed and made availablein the NLP community.While most of these come with word polarity annotations(e.g. positive/negative), attempts at building lexica for finer-grained emotionanalysis (e.g. happiness, sadness) have recently attracted significantattention.Such lexica are often exploited as a building block in the processof developing learning models for which emotion recognition is needed, and/orused as baselines to which compare the performance of the models.In this work,we contribute two new resources to the community: a) an extension of anexisting and widely used emotion lexicon for English;and b) a novel version ofthe lexicon targeting Italian.Furthermore, we show how simple techniques canbe used, both in supervised and unsupervised experimental settings, to boostperformances on datasets and tasks of varying degree of domain-specificity.机器学习论文:DEPECHEMOOD ++:通过简单而强大的技术构建的双语情感词典(DepecheMood++: a Bilingual Emotion Lexicon Built Through Simple Yet  Powerful Techniques) MZz5ewpskcpZC0q8.jpg
URL地址:https://arxiv.org/abs/1810.03660     ----pdf下载地址:https://arxiv.org/pdf/1810.03660    ----机器学习论文:DEPECHEMOOD ++:通过简单而强大的技术构建的双语情感词典(DepecheMood++: a Bilingual Emotion Lexicon Built Through Simple Yet  Powerful Techniques)
回复

使用道具 举报

您需要登录后才可以回帖 登录 | 立即注册

本版积分规则 返回列表 发新帖

tetra当前离线
新手上路

查看:194 | 回复:0

快速回复 返回顶部 返回列表