Weakly-Supervised Deep Learning for Customer Review Sentiment Classification

Authors: Ziyu Guan, Long Chen, Wei Zhao, Yi Zheng, Shulong Tan, Deng Cai

IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on review data obtained from Amazon show the efficacy of our method and its superiority over baseline methods.
Researcher Affiliation Collaboration 1Northwest University 2Xidian University 3Zhejiang University 4Baidu Big Data Laboratory
Pseudocode No The paper describes the network architecture and training steps in narrative text and diagrams (Figure 3), but does not include any formal pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any statement or link regarding the public availability of its source code.
Open Datasets Yes All unlabeled reviews were extracted from the Amazon data product dataset [Mc Auley et al., 2015].
Dataset Splits Yes The labeled dataset was randomly split into training set (50%), validation set (20%) and test set (30%) and we maintain the proportion as shown in Table 1.
Hardware Specification Yes The training is accelerated using GPU (28min for processing 1M triplets on a Nvidia GTX 980ti GPU).
Software Dependencies No The paper mentions software components like word2vec, Liblinear, and Ada Grad but does not provide specific version numbers for them or any other software dependencies.
Experiment Setup Yes We empirically set context vector size to 50, number of filters for each window size to 200, and both hidden layer size and embedding layer size to 300. Hyperbolic tangent is employed as the activation function for all layers. In this paper we set λ = 5.