Text Style Transfer Data

• The classifier in Evaluation is based on keras: https://github.com/fchollet/keras/blob/master/examples/imdb_bidirectional_lstm.py
• The data is available at https://github.com/fuzhenxin/textstyletransferdata
• If there is any problem, please contact Zhenxin Fu (fuzhenxin95@gmail.com)

Evaluation

Evaluation tool is in eval

Preprocess

• put glove embedding in eval/word_emb.
• Please follow run1.sh to put test data.
• The results of converting test data to style 0 is put in style0.txt and style 1 respectively.
• test1 test2 test3 for different mode (autoencoder, style embedding. multi decoder).

Transfer Strength (Classifier)

$ python classifier data        # process data of classifier
$ python classifier train       # train classifier
$ python classifier test test1  # test classifier
                                # test1 is the test result dir
                                # results in test1/embedding/style0_classification.txt ...

Content reservation

$cd eval
$python emb_test.py test1   # test1 is the test result dir
                           # results in test1/embedding/style0_semantics.txt ...

Finally, run python eval.py to show results collection.

Example:

dir_name model_type      transfer_strength content_reservation mixture
================================================================================
test1     embedding8 		0.267 		0.943880306299 	0.208126303212
test1 	embedding4 		0.485 		0.915346000157 	0.317023657029
test1 	embedding 		0.593 		0.896598659955 	0.356930373024
.................

Acknowledgment

Thanks for Fangfang Zhang and Yixin Zhang for helping compose data.