Finetuned BERT model for 13-class classification, without a discourse relation (Expansion.Conjunction). It was introduced in the paper: Automatic Slide Generation Using Discourse Relations and first released in this repository. This model is uncased: it does not make a difference between english and English.

In our proposed method in this paper, we used this model for the classification of discourse relation between the SECOND and THIRD sentence and beyond in summarized sentences. The model is NOT used between the FIRST and SECOND sentences.

This model can classify the relation between the sentence pair of input.

Now we are working on preparing the Model card. Please wait for a few days.

The model trained from bert-large-uncased on the dataset published in the paper: Automatic Prediction of Discourse Connectives.

The dataset to make this model is based on English Wikipedia data and has 20 labels. However, this model will classify into 13 labels. This is because the 20-class data set was restructured to 14 classes to suit our research objective of "automatic slide generation. This distribution is shown below.

This model doesn't contain the discourse relation: Expansion.Conjunction. Because this discourse relation assumes that there is a relation between one previous sentence pair. So it is inappropriate to apply this discourse relation between the first and second sentences.