This paper focuses on how to improve aspect-level opinion mining for online customer reviews. We first propose a novel generative topic model, the Joint Aspect/Sentiment (JAS) model, to jointly extract aspects and asp...This paper focuses on how to improve aspect-level opinion mining for online customer reviews. We first propose a novel generative topic model, the Joint Aspect/Sentiment (JAS) model, to jointly extract aspects and aspect-dependent sentiment lexicons from online customer reviews. An aspect-dependent sentiment lexicon refers to the aspect-specific opinion words along with their aspect-aware sentiment polarities with respect to a specific aspect. We then apply the extracted aspectdependent sentiment lexicons to a series of aspect-level opinion mining tasks, including implicit aspect identification, aspect-based extractive opinion summarization, and aspect-level sentiment classification. Experimental results demonstrate the effectiveness of the JAS model in learning aspectdependent sentiment lexicons and the practical values of the extracted lexicons when applied to these practical tasks.展开更多
面向方面目标的意见词抽取(Target-oriented Opinion Word Extraction,TOWE)是方面情感分析的一个子任务,旨在针对给定句子的特定方面目标识别意见词.在同一句子中,对于不同的目标,模型需要输出不同的抽取结果,TOWE可以给出每个方面目...面向方面目标的意见词抽取(Target-oriented Opinion Word Extraction,TOWE)是方面情感分析的一个子任务,旨在针对给定句子的特定方面目标识别意见词.在同一句子中,对于不同的目标,模型需要输出不同的抽取结果,TOWE可以给出每个方面目标具体的观点信息,从而帮助理解用户情感的来源,细化情感分析任务的粒度.为识别文本中给定方面目标的方面意见词,需要综合考虑上下文语境、方面目标语义信息以及位置信息,并建模候选意见词与方面目标之间的对应关系.本文提出了一个目标语义与位置融合的方面意见词抽取模型(Target-oriented Opinion Word Extraction Based on the Fusion of Aspect Target Semantics and Position model,AP-IOG),模型使用编码器-解码器框架,在方面目标融合编码器中包含3个LSTM模型:向内Inward-LSTM可以充分利用方面目标信息;向外Outward-LSTM可以将方面目标信息很好地编码到上下文中;位置注意力增强的Global-LSTM,可以帮助理解整个句子的全局含义,并且关注到整个句子中方面目标附近的局部信息.这3个LSTM很好地融合了方面目标及其上下文和位置信息,有利于定位针对方面目标的意见词.编码后,将方面目标的上、下文与位置注意力增强的全局上下文进行拼接,传入解码器中.使用TOWE中电脑和餐厅评论领域4个数据集作为实验数据集,实验结果表明,AP-IOG模型明显优于其他方法,在4个数据集上F 1值相比于TOWE的基准模型IOG分别提升了2.23%、2.10%、2.75%以及3.55%.展开更多
基金supported by National Natural Science Foundation of China under Grants No.61232010, No.60903139, No.60933005, No.61202215, No.61100083National 242 Project under Grant No.2011F65China Information Technology Security Evaluation Center Program under Grant No.Z1277
文摘This paper focuses on how to improve aspect-level opinion mining for online customer reviews. We first propose a novel generative topic model, the Joint Aspect/Sentiment (JAS) model, to jointly extract aspects and aspect-dependent sentiment lexicons from online customer reviews. An aspect-dependent sentiment lexicon refers to the aspect-specific opinion words along with their aspect-aware sentiment polarities with respect to a specific aspect. We then apply the extracted aspectdependent sentiment lexicons to a series of aspect-level opinion mining tasks, including implicit aspect identification, aspect-based extractive opinion summarization, and aspect-level sentiment classification. Experimental results demonstrate the effectiveness of the JAS model in learning aspectdependent sentiment lexicons and the practical values of the extracted lexicons when applied to these practical tasks.
文摘面向方面目标的意见词抽取(Target-oriented Opinion Word Extraction,TOWE)是方面情感分析的一个子任务,旨在针对给定句子的特定方面目标识别意见词.在同一句子中,对于不同的目标,模型需要输出不同的抽取结果,TOWE可以给出每个方面目标具体的观点信息,从而帮助理解用户情感的来源,细化情感分析任务的粒度.为识别文本中给定方面目标的方面意见词,需要综合考虑上下文语境、方面目标语义信息以及位置信息,并建模候选意见词与方面目标之间的对应关系.本文提出了一个目标语义与位置融合的方面意见词抽取模型(Target-oriented Opinion Word Extraction Based on the Fusion of Aspect Target Semantics and Position model,AP-IOG),模型使用编码器-解码器框架,在方面目标融合编码器中包含3个LSTM模型:向内Inward-LSTM可以充分利用方面目标信息;向外Outward-LSTM可以将方面目标信息很好地编码到上下文中;位置注意力增强的Global-LSTM,可以帮助理解整个句子的全局含义,并且关注到整个句子中方面目标附近的局部信息.这3个LSTM很好地融合了方面目标及其上下文和位置信息,有利于定位针对方面目标的意见词.编码后,将方面目标的上、下文与位置注意力增强的全局上下文进行拼接,传入解码器中.使用TOWE中电脑和餐厅评论领域4个数据集作为实验数据集,实验结果表明,AP-IOG模型明显优于其他方法,在4个数据集上F 1值相比于TOWE的基准模型IOG分别提升了2.23%、2.10%、2.75%以及3.55%.
