Skip to main content
. 2024 Dec 20;26(12):1114. doi: 10.3390/e26121114
Algorithm 3 Classify Generated Story By Sentence.
  •   1:

    Input: generated_story, bert_classifier

  •   2:

    Output: sentence_classification_results (list of Suitable/Unsuitable)

  •   3:

    Begin

  •   4:

    Split generated_story into individual sentences

  •   5:

    Initialize sentence_classification_results as an empty list

  •   6:

    for each sentence in generated_story do

  •   7:

        Preprocess the sentence

  •   8:

           Tokenize the sentence text

  •   9:

           Convert the sentence to the input format required by BERT

  • 10:

        Feed the preprocessed sentence into the fine-tuned BERT Classifier

  • 11:

        Perform classification

  • 12:

           Obtain prediction probabilities for “suitable” and “unsuitable”

  • 13:

           Set threshold for determining suitability (e.g., 0.5 probability)

  • 14:

        if prediction > threshold for “suitable” then

  • 15:

            Append “Suitable” to sentence_classification_results

  • 16:

        else

  • 17:

            Append “Unsuitable” to sentence_classification_results

  • 18:

        end if

  • 19:

    end for

  • 20:

    Return sentence_classification_results

  • 21:

    End