Supplementary materials of the paper entitled: "Generating Natural Language Requirements via Adversarial Examples in Deep Learning"

Supplementary materials of the paper entitled: "Generating Natural Language Requirements via Adversarial Examples in Deep Learning"

File A: Datasets (.txt)

    A1: Webex

    A2: Zoom

    A3: Teams

    A4: Word

    A5: PowerPoint

    A6: Excel

File B: Python Code (Both ours and baseline)

    B1: adversarial_samples.ipynb

    B2: Baseline.ipynb

FIle C: Result Tables (.xlsx)

    C1: Table of perturbed outputs in Webex 

    C2: Table of perturbed outputs in Zoom

    C3: Table of perturbed outputs in Teams

    C4: Table of perturbed outputs in Word

    C5: Table of perturbed outputs in PowerPoint

    C6: Table of perturbed outputs in Excel (Excel)

File  D: Trend of Adversarial Shifts (Graphs)

    D1: Adversarial shifts of office suit (LSTM)

    D2: Adversarial shifts of video conferencing suit (LSTM)

    D3: Non-Adversarial shifts of office suit (LSTM)

    D4: Non-Adversarial shifts of video conferencing suit(LSTM)

    D5: Adversarial shifts of office suit (GRU)

    D6: Adversarial shifts of video conferencing suit (GRU)

    D7: Non-Adversarial shifts of office suit (GRU)

    D8: Non-Adversarial shifts of video conferencing suit(GRU)

    D9: Adversarial shifts of office suit (Bi-LSTM)

    D10: Adversarial shifts of video conferencing suit (Bi-LSTM)

    D11: Non-Adversarial shifts of office suit (Bi-LSTM)

    D12: Non-Adversarial shifts of video conferencing suit(Bi-LSTM)

File E: Adversarial Examples (.pdf)

    E1: Adversarial vs original in Webex

    E2: Adversarial vs original in Zoom

    E3: Adversarial vs original in Teams

    E4: Adversarial vs original in Word

    E5: Adversarial vs original in PowerPoint

    E6: Adversarial vs original in Excel

File F: Questionnaire