JAIST Repository: A Survey on Deep Learning Models for Aspect-based Sentiment Analysis [課題研究報告書]
2
0
0
全文
(2) A Survey on Deep Learning Models for Aspect-based Sentiment Analysis 1810199. YU Yang. Due to the development of Internet technology, nowadays, our society has entered the Internet era. People around the world can share their lives on the Internet easily. For both buyers and sellers in marketing, these contents (such as Weibo, comments, ¥dots) are the worthiest resources for not only satisfying costumers’ requests but also exploring a new market or exploiting a new product. However, it is challenging to monitor the enormous numbers of comments on the Websites manually. Additionally, different experts have different opinions, and that can easily cause biases in practical work. In recent years, there is a new research direction in Natural Language Processing (NLP) called Sentiment Analysis (or Opinion Mining), which can systematically and automatically analyze text with opinions. This thesis first survey some related knowledge about sentiment analysis and fundamental concepts of deep learning models. Secondly, we investigate the word embedding and some deep learning models, such as Recurrent Neural Network and its variants, Convolutional Neural Network, Attention model, and Transformer. Finally, we analyze the performance of sentiment analysis using those above deep learning models and make our conclusion. As the results showed, among these deep learning models, the Transformer-based sentiment analysis models have achieved the new state-of-the-art accuracy. It also shows the trends of development of NLP, which could contribute further research in this domain. Keywords: sentiment analysis - word representation - deep leaning models..
(3)
関連したドキュメント
Laplacian on circle packing fractals invariant with respect to certain Kleinian groups (i.e., discrete groups of M¨ obius transformations on the Riemann sphere C b = C ∪ {∞}),
Some moment results about the limit of a martingale related to the supercritical branching random walk and perpetuities. Moments for first-passage and last-exit times, the minimum,
She reviews the status of a number of interrelated problems on diameters of graphs, including: (i) degree/diameter problem, (ii) order/degree problem, (iii) given n, D, D 0 ,
The general context for a symmetry- based analysis of pattern formation in equivariant dynamical systems is sym- metric (or equivariant) bifurcation theory.. This is surveyed
The behavior of the derivative of some Kubota- Leopoldt p-adic L-function with trivial zero has a deep relation with some arithmetic Iwasawa module (see [6]).. The second such
— In this paper, we give a brief survey on the fundamental group of the complement of a plane curve and its Alexander polynomial.. We also introduce the notion of
• Using the results of the previous sections, we show the existence of solutions for the inhomogeneous skew Brownian equation (1.1) in Section 5.. We give a first result of
We consider numerical simulations of a compressible fluid in a spherical shell rotating at a constant rotation rate ⌦ about the z-axis.. Entropy is given in units of s, the
