Sarcasm detection on tweets using neural network.
[This repository] perform[s] semantic modelling of sentences using neural networks for the task of sarcasm detection (Ghosh & Veale, 2016).

• nltk (TweetTokenizer)
• Keras
• Tensorflow
• numpy
• scipy
• gensim (if you are using word2vec)
• itertools

git clone [email protected]:AniSkywalker/SarcasmDetection.git
cd SarcasmDetection/src/

If you want to use the pre-trained model, you'll have to download it from Google Drive and save it into /resource/text_model/weights/.

This code is run by the following command:

python sarcasm_detection_model_CNN_LSTM_DNN.py

Please cite the following paper when using this code:

Fracking Sarcasm using Neural Network.
Aniruddha Ghosh and Tony Veale. 7th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis (WASSA 2016). NAACL-HLT. 16th June 2016, San Diego, California, U.S.A.

The supplied input is rated as either 0 meaning non-sarcastic or 1 meaning sarcastic.

If you want to train the model with your own data, you can save your train, development and test data into the /resource/train, /resource/dev and /resource/test folders correspondingly.

The system accepts dataset in the tab separated format — as shown below. An example can be found in /resource/train/train_v1.txt.

id<tab>label<tab>tweet

To run the model with context information and psychological dimensions (using Tensorflow) run:

python sarcasm_context_moods.py

Please cite the following paper when using context information and psychological dimensions:

Magnets for Sarcasm: Making Sarcasm Detection Timely, Contextual and Very Personal
Aniruddha Ghosh and Tony Veale. Conference on Empirical Methods in Natural Language Processing (EMNLP). 7th-11th September, 2017, Copenhagen, Denmark.

• Samples of train, dev, and test files are included for both versions.
• For a test data set, please contact at [email protected]