Interactive Attention Transfer Network for Cross-Domain Sentiment Classification
Authors: Kai Zhang, Hefu Zhang, Qi Liu, Hongke Zhao, Hengshu Zhu, Enhong Chen5773-5780
AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on the Amazon reviews dataset and crowdfunding reviews dataset not only demonstrate the effectiveness and universality of our method, but also give an interpretable way to track the attention information for sentiment. |
| Researcher Affiliation | Collaboration | Anhui Province Key Laboratory of Big Data Analysis and Application, University of Science and Technology of China School of Data Science, University of Science and Technology of China Baidu Talent Intelligence Center, Baidu Inc |
| Pseudocode | No | The paper describes the architecture and components of IATN in detail but does not include any structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not include any explicit statement about making the source code available or provide a link to a code repository for the methodology described. |
| Open Datasets | Yes | For the reliability and authority of experimental results, we use the Amazon reviews dataset, which has been widely used for cross-domain sentiment classification. Meanwhile, we make the necessary pre-processing as follows. First, we choose the reviews data from four domains: Book (B), DVD (D), Electronics (E) and Kitchen appliances (K). Each of the domains contains 6,000 labeled data, in which there are 3,000 positive reviews (higher than 3 stars) and 3,000 negative reviews (lower than 3 stars). Additionally, the dataset also contains lots of unlabeled data. Here we randomly select 8,000 unlabeled reviews from each domain as training data. Table 1 summarizes the statistics of dataset after pre-processing. |
| Dataset Splits | Yes | Table 1: Statistics of datasets after pre-processing. Domains # Train # Test # Unlabel. For all domains: Train 5,000, Test 1,000, Unlabel 8,000. |
| Hardware Specification | No | The paper does not provide specific hardware details such as GPU/CPU models, processor types, or memory amounts used for running its experiments. |
| Software Dependencies | No | The paper mentions 'word2vec' but does not specify its version or list other software dependencies with version numbers needed for replication. |
| Experiment Setup | Yes | In our experiments, all word embeddings from sentences and aspects are initialized as 200-dimension vectors by word2vec (Goldberg and Levy 2014). The dimensions of word embeddings, attention vectors and LSTM hidden states are set to 200, 64 and 64 respectively. All weight matrices are randomly initialized by a uniform distribution U( 0.01, 0.01), and all biases are set to zeros. For the performance of IATN, we finally set the coefficient of l2 normalization, the learning rate and the dropout rate as 10 4, 10 3 and 0.25. |