🥸 Text Anonymization model

Use the default Text Anonymization model

Need a general-purpose Text Anonymization model? You can use Artifex's default Text Anonymization model, which is trained to recognize and remove five Personal Identifiable Information (PII) types out-of-the-box:

• PERSON
• LOCATION
• DATE
• ADDRESS
• PHONE_NUMBER

from artifex import Artifex

ta = Artifex().text_anonymization

print(ta("John Doe lives at 123 Main St, New York. His phone number is (555) 123-4567."))
# >>> ["[MASKED] lives at [MASKED]. His phone number is [MASKED]."]

Learn more about the default Text Anonymization model on our Text Anonymization HF model page.

Create & use a custom Text Anonymization model

Do you want to tailor the model to your specific domain for better results? Fine-tune your own Text Anonymization model, use it locally on CPU and keep it forever:

from artifex import Artifex

ta = Artifex().text_anonymization

model_output_path = "./output_model/"

ta.train(
    domain="medical documents", # change to your desired domain
    output_path=model_output_path
)

ta.load(model_output_path)
print(ta("The patient John Doe visited New York on 12th March 2023 at 10:30 AM."))

# >>> ["The patient [MASKED] visited [MASKED] on [MASKED] at [MASKED]."]