Skip to content
Deriving Machine Attention from Human Rationales
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

Deriving Machine Attention from Human Rationales

This repo contains the code and data of the following paper:

Deriving Machine Attention from Human Rationales. Yujia Bao, Shiyu Chang, Mo Yu and Regina Barzilay. EMNLP 2018.

If you find this work useful and use it on your own research, please cite our paper.

  title={Deriving Machine Attention from Human Rationales},
  author={Bao, Yujia and Chang, Shiyu and Yu, Mo and Barzilay, Regina},
  journal={arXiv preprint arXiv:1808.09367},


The R2A model first learns to map binary rationales into continuous attention scores on the source tasks. Then the trained R2A model is used to predict how attention should look like based on human-annotated rationales for the low-resource target task. Finally, we train a target classifier under the supervision of both the annotated labels and the R2A-generated attention. The following figure illustrates our learning pipeline.



Instructions to run the code are provided within each directory.

  • Directory r2a contains the source code and pre-trained models for our R2A model.
  • Directory rationalization contains the code we used for automatic rationale generation.



The original raw dataset can be found at: beer review, hotel review.

We provide the processed data (together with the machine-generated rationales) that we used for all our experiments at Important Note: this data is for research-purpose only.


  1. Unzip to the root directory of this repo.
  2. There are three directories under the directory data, named as source, target and oracle.
    • source includes all source data files. Each data file is a tsv file that contains the following fields: task name, label, text (tokenized and separated by space), rationale label (a sequence of binary integer separated by space).

      Task #train (file) #dev (file)
      Beer look 43,351 (beer0.train) 10,170 (
      Beer aroma 39,825 (beer1.train) 8,772 (
      Beer palate 30,041 (beer2.train) 7,152 (
    • oracle contains the data used to derive the oracle attention. The data format is the same as the one in source.

      Task #train (file) #dev (file)
      Beer look 32,276 (beer0.train) 6392 (
      Beer aroma 28,984 (beer1.train) 5,720 (
      Beer palate 25,748 (beer2.train) 4,994 (
      Hotel location 14,472 (hotel_Location.train) 1,813 (
      Hotel cleanliness 150,098 (hotel_Cleanliness.train) 18,764 (
      Hotel service 101,484 (hotel_Service.train) 12,689 (
    • target contains the data for the target tasks.

      • hotel_unlabeled.train, unlabeled data file. Each row is a hotel review. Used for training the domain-invariant encoder of our R2A model.

      • *.dev, *.test: target development and test set. The data format is the same as the one in source.

        Task #dev (file) #test (file)
        Beer look 200 ( 4014 (beer0.test)
        Beer aroma 200 ( 4212 (beer1.test)
        Beer palate 200 ( 3804 (beer2.test)
        Hotel location 200 ( 1808 (hotel_Location.test)
        Hotel cleanliness 200 ( 12684 (hotel_Cleanliness.test)
        Hotel service 200 ( 18762 (hotel_Service.test)
      • *.train: target training set. Each data file (except and hotel_unlabeled.train) is a tsv file that contains the following fields: 1) task name, 2) label, 3) text (tokenized and separated by space), 4) rationale label (a sequence of binary integer), 5) R2A-generated attention (a sequence of float), 6) oracle attention (a sequence of float), frequency of a word being highlighted as rationale (a sequence of float).

        • beer0.train (beer look), beer1.train (beer aroma), beer2.train (beer palate), hotel_Location.train, hotel_Cleanliness.train, hotel_Service.train: Each data file consists of 200 labeled examples with human annotated rationales. The entries for R2A-generated attention and oracle attention are all zero.
        • *.pred_att.gold_att.train: the file contains R2A-generated attention and the oracle attention from pretrained models.


  • PyTorch 0.4.1
  • numpy 1.15.1
  • torchtext 0.2.1
  • termcolor 1.1.0
  • tqdm 4.24.0
  • scikit-learn 0.19.2
  • spacy 2.0.12
  • colored 1.3.5
You can’t perform that action at this time.