Sentic-GCN icon indicating copy to clipboard operation
Sentic-GCN copied to clipboard

[KBS] Aspect-based sentiment analysis via affective knowledge enhanced graph convolutional networks

Results 15 Sentic-GCN issues
Sort by recently updated
recently updated
newest added

Hi! I wonder whether a newer version of senticnet, for instance senticnet-7, could be used as an alternative of the senticnet-5 in the model? Thanks!

Hi, Could you please tell me what is the difference between using a knowledge base like senticnet and a normal sentiment lexicon that provide words with their sentiment plarities? Many...

FileNotFoundError: [Errno 2] No such file or directory: './datasets/acl-14-short-data/train.raw.graph'这个错误如何改,谢谢

loading word vectors ... Traceback (most recent call last): File "D:\sherif\B thesis\sarcasm paper\import papers\Sentic-GCN-main\ABSA-Detection\infer.py", line 124, in inf = Inferer(opt) File "D:\sherif\B thesis\sarcasm paper\import papers\Sentic-GCN-main\ABSA-Detection\infer.py", line 55, in __init__ embedding_matrix...

您好,作者。我在研读代码的过程中,发现generate_sentic_dependency_graph.py中的生成情感增强邻接矩阵算法与论文中的算法1并不相同。因此我在代码中改正成论文中的算法,但似乎造成了梯度爆炸。以下是我改正后的代码和train图,不知是否是我的理解有偏差,希望得到您的回答! ![Snipaste_2023-08-02_10-06-18](https://github.com/BinLiang-NLP/Sentic-GCN/assets/39265702/24b16fdb-55d2-47ed-84b0-b1b690c8623b) ![Snipaste_2023-08-02_10-11-39](https://github.com/BinLiang-NLP/Sentic-GCN/assets/39265702/b31f74a5-c9ca-429f-b790-97711d295227) ![Snipaste_2023-08-02_10-11-47](https://github.com/BinLiang-NLP/Sentic-GCN/assets/39265702/4c52166a-cfb8-42e1-88a8-2d18530fa8e6)

Hello, what's the difference with senticNet20, 30, 40 ...

First of all, thank you for releasing the program on your paper. What I'm curious about is that SpaCy divides sentences into word units, but BERT divides them into WordPiece...

您好!阅读了您的论文之后,对于情感增强依存图的意义还不是很理解,添加了情感知识后的邻接矩阵中的数值有什么含义,在图卷积中是不是把这个数值看作两个单词间的相关性。以下面的例子为例: ![例子](https://user-images.githubusercontent.com/56523651/204211700-d7c50705-e3a6-4ba1-a069-aa6d5492d5f8.jpg) 这句话中的方面词分别是food和service,根据依存树得到的邻接矩阵中food和delicious,以及service和terrible的边的权重为1,在senticnet词典中,food、delicious、serviceh和terrible对应的情感值分别为0.054、0.106、0.84和-0.90,加入情感知识后,food和delicious,以及service和terrible的边的权重变为2.16,和1.94,请问这个数值的含义是什么,希望能尽快得到您的回复。

您好,我用您的代码跑中文数据集出现了以下错误,请问您知道解决方案吗? Traceback (most recent call last): File "E:/Sentic-GCN-main/Sentic-GCN-main/train_bert.py", line 281, in main() File "E:/Sentic-GCN-main/Sentic-GCN-main/train_bert.py", line 276, in main ins = Instructor(opt) File "E:/Sentic-GCN-main/Sentic-GCN-main/train_bert.py", line 53, in __init__ self.trainset = ABSADataset(opt.dataset_file['train'],...