A Unified Generative Framework for Aspect-Based Sentiment Analysis论文速看

)

总括

ABSA旨在识别： aspect terms, their corresponding sentiment polarities, and the opinion terms。文章总结了7类ABSA的子任务，看这些子任务，我们可以观察到：input不一样，output也不一样，任务类型也不一样。所以现有的一些研究都是只做1-3个子任务的，还未有一个方法去一下子解决7类子任务。本文就是做了这样一件事！此paper把这7个subtask都变成了一个 unified generative formulation.

Introduction

ABSA是细粒度的情感分析任务，旨在识别a,s,o.( aspect terms, sentiment polarities, and the opinion terms。)

1. 看例子 ，
   黄色字体是aspect term ,粉色字是opinion term,情感都是positive。
2. 7subtask总结
   
   可以看到，有不同的输入输出以及任务类型。
   正是这些divergences才导致make it difficult to solve all subtask in a unified framework.
3. recently work the existing methods can hardly solve all the subtasks by a unified framework without relying on the sub-models or changing the model structure to adapt to all ABSA subtasks。所以我们提出的统一生成框架去解决这7类任务很有必要。
4. 贡献
   • We formulate both the extraction task and classification task of ABSA into a unified index generation problem. Unlike previous unified models,our method needs not to design specific decoders for different output types.
   • With our re-formulation, all ABSA subtasks
   can be solved in sequence-to-sequence framework, which is easy-to-implement and can be built on the pre-trained models, such as BART.
   • We conduct extensive experiments on four public datasets, and each dataset contains a subset of all ABSA subtasks. To the best of our knowledge, it is the first work to evaluate a model on all ABSA tasks.
   • The experimental results show that our proposed ramework significantly outperforms recent SOTA methods.

Methodology

• Task Formulation
  -For example, o s, a erepresent the start index of an opinion term o and the end index of an aspect term a. We use the s p to denote the index of sentiment polarity class. The target sequence for each subtask is as follows:
  
  
  
  把这七类都转换成目标序列生成
  

• Our model

• 
  

框架的整体结构，图底下有详细介绍。在训练阶段，我们使用了 teacher
forcing 来训练 model and the negative loglikelihood to optimize the model. Moreover, during the inference, we use the beam search to get
the target sequence Y in an autoregressive manner.

实验

dataset

baseline

main result

Framework Analysis

为了证明我们的生成框架能够适应ABSA生成任务，我们进行了测试，定义了三种错误“invalid token”,“invalid size”,“invalid order”

实验证实错误都很少，所以说我们这个生成框架基本上没有什么问题。

conclusion

论文不是很长，也就七八页，详细内容请看原文。（上面放的都是我之前做的PPT）本人也是第一次写博客，有不足之处请多包含。有错误也请批评指正。