End-to-end aspect category sentiment analysis based on type graph convolutional networks

For the existing aspect category sentiment analysis research,most of the aspects are given for sentiment extraction,and this pipeline method is prone to error accumulation,and the use of graph convolutional neural network for aspect category sentiment analysis does not fully utilize the dependency type information between words,so it cannot enhance feature extraction.This paper proposes an end-to-end aspect category sentiment analysis(ETESA)model based on type graph convolutional networks.The model uses the bidirectional encoder representation from transformers(BERT)pretraining model to obtain aspect categories and word vectors containing contextual dynamic semantic information,which can solve the problem of polysemy;when using graph convolutional network(GCN)for feature extraction,the fusion operation of word vectors and initialization tensor of dependency types can obtain the importance values of different dependency types and enhance the text feature representation;by transforming aspect category and sentiment pair extraction into multiple single-label classification problems,aspect category and sentiment can be extracted simultaneously in an end-to-end way and solve the problem of error accumulation.Experiments are tested on three public datasets,and the results show that the ETESA model can achieve higher Precision,Recall and F1 value,proving the effectiveness of the model.

Improve Chinese Aspect Sentiment Quadruplet Prediction via Instruction Learning Based on Large Generate Models

Aspect-Based Sentiment Analysis(ABSA)is a fundamental area of research in Natural Language Processing(NLP).Within ABSA,Aspect Sentiment Quad Prediction(ASQP)aims to accurately identify sentiment quadruplets in target sentences,including aspect terms,aspect categories,corresponding opinion terms,and sentiment polarity.However,most existing research has focused on English datasets.Consequently,while ASQP has seen significant progress in English,the Chinese ASQP task has remained relatively stagnant.Drawing inspiration from methods applied to English ASQP,we propose Chinese generation templates and employ prompt-based instruction learning to enhance the model’s understanding of the task,ultimately improving ASQP performance in the Chinese context.Ultimately,under the same pre-training model configuration,our approach achieved a 5.79%improvement in the F1 score compared to the previously leading method.Furthermore,when utilizing a larger model with reduced training parameters,the F1 score demonstrated an 8.14%enhancement.Additionally,we suggest a novel evaluation metric based on the characteristics of generative models,better-reflecting model generalization.Experimental results validate the effectiveness of our approach.

Fusing Syntactic Structure Information and Lexical Semantic Information for End-to-End Aspect-Based Sentiment Analysis

The aspect-based sentiment analysis(ABSA)consists of two subtasksaspect term extraction and aspect sentiment prediction.Most methods conduct the ABSA task by handling the subtasks in a pipeline manner,whereby problems in performance and real application emerge.In this study,we propose an end-to-end ABSA model,namely,SSi-LSi,which fuses the syntactic structure information and the lexical semantic information,to address the limitation that existing end-to-end methods do not fully exploit the text information.Through two network branches,the model extracts syntactic structure information and lexical semantic information,which integrates the part of speech,sememes,and context,respectively.Then,on the basis of an attention mechanism,the model further realizes the fusion of the syntactic structure information and the lexical semantic information to obtain higher quality ABSA results,in which way the text information is fully used.Subsequent experiments demonstrate that the SSi-LSi model has certain advantages in using different text information.

Syntax-Based Aspect Sentiment Quad Prediction by Dual Modules Neural Network for Chinese Comments

Aspect-Based Sentiment Analysis(ABSA)is one of the essential research in the field of Natural Language Processing(NLP),of which Aspect Sentiment Quad Prediction(ASQP)is a novel and complete subtask.ASQP aims to accurately recognize the sentiment quad in the target sentence,which includes the aspect term,the aspect category,the corresponding opinion term,and the sentiment polarity of opinion.Nevertheless,existing approaches lack knowledge of the sentence’s syntax,so despite recent innovations in ASQP,it is poor for complex cyber comment processing.Also,most research has focused on processing English text,and ASQP for Chinese text is almost non-existent.Chinese usage is more casual than English,and individual characters contain more information.We propose a novel syntactically enhanced neural network framework inspired by syntax knowledge enhancement strategies in other NLP studies.In this framework,part of speech(POS)and dependency trees are input to the model as auxiliary information to strengthen its cognition of Chinese text structure.Besides,we design a relation extraction module,which provides a bridge for the overall extraction of the framework.A comparison of the designed experiments reveals that our proposed strategy outperforms the previous studies on the key metric F1.Further experiments demonstrate that the auxiliary information added to the framework improves the final performance in different ways.

A Multitask Multiview Neural Network for End-to-End Aspect-Based Sentiment Analysis

The aspect-based sentiment analysis(ABSA) consists of two subtasks—aspect term extraction and aspect sentiment prediction. Existing methods deal with both subtasks one by one in a pipeline manner, in which there lies some problems in performance and real application. This study investigates the end-to-end ABSA and proposes a novel multitask multiview network(MTMVN) architecture. Specifically, the architecture takes the unified ABSA as the main task with the two subtasks as auxiliary tasks. Meanwhile, the representation obtained from the branch network of the main task is regarded as the global view, whereas the representations of the two subtasks are considered two local views with different emphases. Through multitask learning, the main task can be facilitated by additional accurate aspect boundary information and sentiment polarity information. By enhancing the correlations between the views under the idea of multiview learning, the representation of the global view can be optimized to improve the overall performance of the model. The experimental results on three benchmark datasets show that the proposed method exceeds the existing pipeline methods and end-to-end methods, proving the superiority of our MTMVN architecture.

基于方面-词性感知的方面级情感分析

方面级情感分析是自然语言处理的研究热点之一,其任务目的是预测句子中给定方面的情感极性。目前已有研究大多忽略了方面词和特定词性单词在过滤情感极性相关上下文语义信息和理解上下文语法信息中的作用。为此,提出一种基于方面-词性感知的图卷积网络ASP_POSGCN。采用双向长短期记忆网络建模上下文和词性信息,经由门控机制筛选方面词相关上下文语义信息,再使用词性信息隐藏层状态进一步过滤;同时设计方面-词性感知矩阵算法,根据不同词性单词对方面词情感极性的贡献重构句子原始依存关系以获取重构依存句法图,将原始依存句法图和重构依存句法图应用于双通道图卷积网络和多图感知机制;最后,使用过滤后的上下文语义信息与双通道图卷积网络的输出计算注意力得到最终分类表示。实验结果表明,该模型在Twitter、Laptop14、Restaurant14和Restaurant164个公开数据集上的准确率分别为74.57%、79.15%、83.84%、91.23%,F1值分别为72.59%、75.76%、77.00%、77.11%,与传统方面级情感分析基准模型相比均有提升,有助于方面级的情感极性分类。

Aspect-level sentiment analysis based on semantic heterogeneous graph convolutional network

The deep learning methods based on syntactic dependency tree have achieved great success on Aspect-based Sentiment Analysis(ABSA).However,the accuracy of the dependency parser cannot be determined,which may keep aspect words away from its related opinion words in a dependency tree.Moreover,few models incorporate external affective knowledge for ABSA.Based on this,we propose a novel architecture to tackle the above two limitations,while fills up the gap in applying heterogeneous graphs convolution network to ABSA.Specially,we employ affective knowledge as an sentiment node to augment the representation of words.Then,linking sentiment node which have different attributes with word node through a specific edge to form a heterogeneous graph based on dependency tree.Finally,we design a multi-level semantic heterogeneous graph convolution network(Semantic-HGCN)to encode the heterogeneous graph for sentiment prediction.Extensive experiments are conducted on the datasets SemEval 2014 Task 4,SemEval 2015 task 12,SemEval 2016 task 5 and ACL 14 Twitter.The experimental results show that our method achieves the state-of-the-art performance.

基于ABSA方法的移动政务用户情感分析

移动政务APP作为各级政府服务群众的重要渠道,其在线评论的情感倾向会对用户的线上政务满意度产生重要影响。为了对当前移动政务APP用户评论进行细粒度情感分析,文章基于ABSA方法进行移动政务用户评论的情感倾向性测度:运用LDA主题模型进行隐式方面主题的抽取,并结合期望确认理论与信息系统成功模型构建移动政务用户需求模型;同时,选择BERT模型对实体词进行情感倾向概率判定,进而通过多维度间的情感倾向匹配实现方面级的情感强度测量。研究发现,通过该方法可以挖掘移动政务用户的需求重点与情感状态,当前用户对服务质量呈现积极的情感状态,而对系统质量与信息质量的情感评价则较为负面。

基于BERT与注意力机制的方面级隐式情感分析模型

在方面级情感文本中存在部分不含情感词的评论句,对其情感的研究被称为方面级隐式情感分析.现有分析模型在预训练过程中可能会丢失与方面词相关的上下文信息,并且不能准确提取上下文中深层特征,本文首先构造了方面词感知BERT预训练模型,通过将方面词引入到基础BERT的输入嵌入结构中,生成与方面词信息相关的词向量;然后构造了语境感知注意力机制,对由编码层得到的深层隐藏向量,将其中的语义和句法信息引入到注意力权重计算过程,使注意力机制能更加准确地分配权重到与方面词相关的上下文.对比实验结果表明,本文模型的效果优于基线模型.

融合匹配长短时记忆网络和语法距离的方面级情感分析模型

针对现阶段方面级情感分析(ABSA)存在的方面词与不相关上下文错误匹配以及缺乏语法层面特征的问题,提出一种融合匹配长短时记忆网络(mLSTM)和语法距离的ABSA模型mLSTM-GCN。首先,逐词计算方面词与上下文的关联性,并将得到的注意力权重与上下文表示融合作为mLSTM的输入,从而得到与方面词关联度更高的上下文表示;然后,引入语法距离以获得与方面词语法关联度更高的上下文,从而获取更多的上下文特征来指导方面词的建模,并通过方面掩盖层得到方面表示;最后,结合位置权重、上下文表示以及方面表示来进行信息交互,从而获取用于情感分析的特征。在Twitter、REST14和LAP14数据集上的实验结果表明,相较于特定方面的图卷积网络(ASGCN),mLSTM-GCN的准确率分别提升1.32、2.50和1.63个百分点,宏平均F1分别提升2.52、2.19和1.64个百分点。可见,mLSTM-GCN能够有效降低方面词与不相关上下文错误匹配的概率,提升分类效果。

融合多窗口局部信息的方面级情感分析模型

针对目前方面级情感分析(ABSA)模型过多依赖关系较为稀疏的句法依赖树学习特征表示,导致模型学习局部信息能力不足的问题,提出了一种融合多窗口局部信息的ABSA模型MWGAT(combining Multi-Window local information and Graph ATtention network)。首先,通过多窗口局部特征学习机制学习局部上下文特征,并挖掘文本包含的潜在局部信息;其次,采用能够较好理解依赖树的图注意力网络(GAT)学习句法依赖树所表示的语法结构信息,并生成语法感知的上下文特征;最后,将这两种表示不同语义信息的特征融合,形成既包含句法依赖树的语法信息又包含局部信息的特征表示,从而便于分类器高效判别方面词的情感极性。在Restaurant、Laptop和Twitter这3个公开数据集上进行实验,结果表明与结合了句法依赖树的T-GCN(Type-aware Graph Convolutional Network)模型相比,所提模型的Macro-F1分数分别提高了2.48%、2.37%和0.32%。可见,所提模型能够有效挖掘潜在的局部信息,并更为精确地预测方面词的情感极性。

基于虚拟依存关系与知识增强的方面级情感分析

借助句法依赖信息和外部知识的图神经网络近年来成为方面级情感分析领域的一个研究热点,但是现有研究存在语法信息提取不充分和利用不合理等问题,同时未考虑增强文本方面词与意见词等关键节点的背景知识。此外,基于注意力机制的方法没有建立方面词与上下文词的语法信息交互,导致方面词错误地关注到与其语法无关的上下文词信息。提出一种基于虚拟依存关系与双知识增强的多交互图卷积网络模型。对方面词内每个单词构建依赖树,依据虚拟依存关系进行加权构造虚拟依存图,依据外部情感知识构造情感依存图,使用双通道图卷积神经网络处理虚拟依存图与情感依存图并进行融合,对融合后特定方面的特征表示进行语义和语法双交互。利用概念知识增强特定方面后的特征表示与上下文表示并进行知识注意力交互,对多种不同的增强表示进行融合从而实现不同表示间的共享与互补。实验结果表明,相较于经典的图卷积网络模型ASGCN,该模型在Rest15和Rest16数据集上的F1值分别提升4.71和8.57个百分点,具有较好的情感分类性能。

嵌入不同邻域表征的方面级情感分析模型

方面级情感分析(ABSA)任务旨在识别特定方面的情感极性,然而现有的相关模型对结构不定的自然语句缺少对方面词上下文的短距离约束,且容易忽略句法关系,因而难以准确判定方面的情感极性。针对上述问题,提出嵌入不同邻域表征(EDNR)的ABSA模型。在该模型中,在获得句子语序信息的基础上,采用近邻策略并结合卷积神经网络(CNN)获取方面的邻域信息,减少较远无关信息对模型的影响;同时,引入语句的语法信息,增加单词之间的依赖关系;将上述两种特征融合后,使用Mask与注意力机制来特别关注方面信息,减少无用信息对情感分析模型的干扰。此外,为评价上下文和语法信息对情感极性的影响程度,提出一个信息评估系数。在5个公共数据集上进行实验的结果表明,与情感分析模型聚合图卷积网络-最大值函数(AGCN-MAX)相比,EDNR模型在数据集14Lap上的正确率和F1值分别提升了2.47和2.83个百分点。由此可见,EDNR模型可以有效捕获情感特征,提高分类性能。

A Semi-Supervised Approach for Aspect Category Detection and Aspect Term Extraction from Opinionated Text

The Internet has become one of the significant sources for sharing information and expressing users’opinions about products and their interests with the associated aspects.It is essential to learn about product reviews;however,to react to such reviews,extracting aspects of the entity to which these reviews belong is equally important.Aspect-based Sentiment Analysis(ABSA)refers to aspects extracted from an opinionated text.The literature proposes different approaches for ABSA;however,most research is focused on supervised approaches,which require labeled datasets with manual sentiment polarity labeling and aspect tagging.This study proposes a semisupervised approach with minimal human supervision to extract aspect terms by detecting the aspect categories.Hence,the study deals with two main sub-tasks in ABSA,named Aspect Category Detection(ACD)and Aspect Term Extraction(ATE).In the first sub-task,aspects categories are extracted using topic modeling and filtered by an oracle further,and it is fed to zero-shot learning as the prompts and the augmented text.The predicted categories are the input to find similar phrases curated with extracting meaningful phrases(e.g.,Nouns,Proper Nouns,NER(Named Entity Recognition)entities)to detect the aspect terms.The study sets a baseline accuracy for two main sub-tasks in ABSA on the Multi-Aspect Multi-Sentiment(MAMS)dataset along with SemEval-2014 Task 4 subtask 1 to show that the proposed approach helps detect aspect terms via aspect categories.

融合多通道图卷积网络的方面级情感分析模型

方面级情感分析是一项细粒度的情感分析任务,旨在对给定文本中多个特定方面进行情感极性分析。大多数基于句法分析的模型严重依赖依存树的单一解析结果,由于这种结构仅揭示2个单词之间的依存关系,因此无法捕捉方面词与上下文之间的联系。此外,大部分研究忽略了依存解析中的关系类型,可能丢失重要特征信息。提出一种基于多通道的图卷积网络(MCGCN)模型,主要包含成分图模块、依存图模块(DepGCN)以及自注意力机制,旨在挖掘句子结构中丰富的句法和语法信息。此外,为减少依存解析错误,在DepGCN中引入依存关系类型和依存概率以获取更精准的语法表示。利用预训练模型进行词嵌入得到文本初始化向量,分别输入到融合不同句法分析的图卷积网络进行不同节点的学习,利用自注意力机制能有效地对方面词进行全局的语义信息关联。最后将3个模块的输出通过仿射交互层进行交互学习,进而提升模型的表征能力。实验结果表明,MCGCN模型在Laptop数据集上的ACC值和F1值分别为78.48%和75.03%,在Twitter数据集上的ACC值和F1值分别为75.92%和74.53%,相较于基准模型均有提升,在3个数据集上的F1值都有1%~3%的提升。

面向双注意力网络的特定方面情感分析模型

特定方面情感分析已经成为自然语言处理领域的研究热点,其通过学习文本上下文的信息判别文本中特定方面的情感极性,可以更加有效地帮助人们了解用户对不同方面的情感表达.当前,将注意力机制和神经网络相结合的模型在解决特定方面情感分析任务时大多仅考虑单一层面的注意力信息,并且卷积神经网络无法获取全局结构信息、循环神经网络训练时间过长且单词间的依赖程度随着距离增加而逐渐减弱.针对上述问题,提出一种面向双注意力网络的特定方面情感分析(dual-attention networks for aspect-level sentiment analysis,DANSA)模型.首先,引入多头注意力机制,通过对输入进行多次不同的线性变换操作,获取更全面的注意力信息,同时,多头注意力机制可以实现并行化计算,保证了DANSA的训练速度.其次,DANSA引入自注意力机制,通过计算输入中每个单词与其他所有单词的注意力得分获取全局结构信息,并且单词间的依赖程度不会受到时间和句子长度的影响.最后,融合上下文自注意力信息与特定方面单词注意力信息,共同作为特定方面情感预测的依据,最终实现特定方面情感极性的预测.相比结合注意力机制的神经网络,DANSA弥补了注意力信息单一问题,不仅可以有效获取全局结构信息,还能够实现并行化计算,大大降低了训练时间.在SemEval2014数据集和Twitter数据集上进行实验,DANSA获得了更好的分类效果,进一步证明了DANSA的有效性.

基于BLSTM与方面注意力模块的情感分类方法

基于方面的情感分析已广泛应用于文本信息挖掘,但在句子情感极性模糊或包含多个不同方面情感极性时难以准确提取特征信息,削弱了情感极性分类效果。为解决该问题,提出一种结合双向长短记忆网络和方面注意力模块的情感分类方法。利用多个方面注意力模块同时对不同方面进行独立训练,使每个方面信息与注意力操作互不影响,各自进行注意力参数的学习与调整,以充分提取特定方面的隐藏信息,从而更准确地识别不同方面的情感极性。在SemEval数据集上的实验结果表明,该方法相对现有的基准情感分析方法,可有效提升分类精确率、查全率与F1值,优化情感分类效果。

基于图神经网络的方面级情感分析

目前基于循环神经网络和注意力机制的方面级情感分析模型缺乏解释相关句法约束和远程单词依赖关系。针对该问题提出结合句子依存树和单词序列信息建立句子关系图模型。首先将句子表示为图,单词作为图的节点,依存句法树的边和单词序列作为图的边;然后提出邻接矩阵标记方案对句子关系图进行标记;最后利用图神经网络实现节点和边的分类任务。该模型在SemEval2014任务中的restaurant和laptop两个数据集上进行实验,在两个数据集上F 1值提升了5%左右。实验结果表明,将句子转换成图利用图神经网络对句子进行方面级情感分析是有益的。

基于BERT与记忆网络的长文本方面级情感分析

目前在方面级情感分析(ABSA)方法中,利用上下文或方面短语的平均值来计算方面短语或上下文之间注意力得分的方法往往会产生较大的信息损失,导致模型在长文本分类上的性能降低。为此,研究了一种建立在BERT表示上的记忆网络模型,BDMN。首先,把句子构造成多[CLS]的Token嵌入形式,然后,从BERT输出中获取到各短句的初始化向量,将其作为记忆体与方面短语的[CLS]向量进行充分注意力交互,最后,拼接输出注意力层所有短句的方面短语表示作为最终的情感分类特征向量。在AI Challenger 2018细粒度情感分析数据集上进行了模型评估,与目前的主流模型相比,其效果达到了最好。

结合语法增强与噪声削减的方面级情感分析模型

方面级情感分析的主要目标是判断句子中给定方面词的情感极性。最近的研究主要采用依存句法信息,隐式地关联方面词与目标词之间的情感交互信息。但结合依存句法信息方法,缺乏识别以方面词为中心的局部上下文信息。此外,对繁杂的语法信息进行等价建模,会引入损害模型性能的噪声。针对以往研究中存在的问题,提出了结合语法增强与噪声削减的神经网络模型。该方法在依存句法信息基础上融合了成分信息,使模型不仅能关注词与词之间的全局依赖信息,同时能关注以方面词为中心的局部依赖信息。同时,为了降低语法信息的噪声干扰,模型以依存句法树的距离信息为依据,弱化了远距离的噪声干扰。最后,模型在四个基准数据集上进行了实验,并在所有数据集上的性能都优于基线模型。