LS-Tree: Model Interpretation When the Data Are Linguistic

Authors: Jianbo Chen, Michael Jordan3454-3461

AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We carry out experiments to analyze the performance of four different models: Bag of Words (Bo W), Word-based Convolutional Neural Network (CNN) (Kim 2014), bidirectional Long Short-Term Memory network (LSTM) (Hochreiter and Schmidhuber 1997), and Bidirectional Encoder Representations from Transformers (BERT) (Devlin et al. 2018), across three sentiment data sets of different sizes: Stanford Sentiment Treebank (SST) (Socher et al. 2013), IMDB Movie reviews (Maas et al. 2011) and Yelp reviews Polarity (Zhang, Zhao, and Le Cun 2015).
Researcher Affiliation Academia Jianbo Chen University of California, Berkeley jianbochen@berkeley.edu Michael I. Jordan University of California, Berkeley jordan@cs.berkeley.edu
Pseudocode Yes Algorithm 1 LS-Tree Interaction Detection and Algorithm 2 Recursion are present.
Open Source Code Yes The code for replicating the experiments is available online at https://github.com/Jianbo-Lab/LS-Tree.
Open Datasets Yes Stanford Sentiment Treebank (SST) (Socher et al. 2013), IMDB Movie reviews (Maas et al. 2011) and Yelp reviews Polarity (Zhang, Zhao, and Le Cun 2015).
Dataset Splits No Table 1 lists 'Train Size' and 'Test Size' for the datasets, but no explicit validation set split is provided for the experiments conducted in the paper.
Hardware Specification No The paper does not provide specific details on the hardware used to run the experiments (e.g., GPU or CPU models).
Software Dependencies No The paper mentions software components like 'GloVe word embedding' and 'Stanford constituency parser' but does not provide specific version numbers for any software dependencies.
Experiment Setup Yes The CNN is composed of three 100-dimensional convolutional 1D layers with 3, 4 and 5 kernels respectively, concatenated and fed into a max-pooling layer followed by a hidden dense layer. The LSTM uses a bidirectional LSTM layer with 128 units for each direction.