Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Renew Phishing Classifier #38644

Open
wants to merge 2 commits into
base: master
Choose a base branch
from
+21 −8
Open

Renew Phishing Classifier #38644

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1363bc5
cherry-pick
jlevypaloalto Feb 13, 2025
ad1bfc3
remove file
jlevypaloalto Feb 13, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
29 changes: 21 additions & 8 deletions Packs/Base/Scripts/DBotPredictPhishingWords/DBotPredictPhishingWords.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,19 @@
UNKNOWN_MODEL_TYPE = 'UNKNOWN_MODEL_TYPE'


class TEMP_RES:
data = {}


def log_io(func, loc):
def new_func(*x, **y):
TEMP_RES.data[f'{loc}-{func.__name__}'] = {'input': (x, y)}
res = func(*x, **y)
TEMP_RES.data[f'{loc}-{func.__name__}']['output'] = res
return res
return new_func


def OrderedSet(iterable):
return list(dict.fromkeys(iterable))

Expand Down Expand Up @@ -130,13 +143,13 @@ def predict_batch_incidents_light_output(email_subject, email_body, phishing_mod
batch_predictions = []
for input_text in preprocessed_text_list:
incident_res = {'Label': -1, 'Probability': -1, 'Error': ''}
filtered_text, filtered_text_number_of_words = phishing_model.filter_model_words(input_text)
filtered_text, filtered_text_number_of_words = log_io(phishing_model.filter_model_words, 'batch')(input_text)
if filtered_text_number_of_words == 0:
incident_res['Error'] = "The model does not contain any of the input text words"
elif filtered_text_number_of_words < min_text_length:
incident_res['Error'] = "The model contains fewer than %d words" % min_text_length
else:
pred = phishing_model.predict(input_text)
pred = log_io(phishing_model.predict, 'batch')(input_text)
incident_res['Label'] = pred[0]
prob = pred[1]
if isinstance(prob, np.floating):
Expand All @@ -156,13 +169,14 @@ def predict_single_incident_full_output(email_subject, email_body, is_return_err
word_threshold):
text = f"{email_subject} \n{email_body}"
input_text, words_to_token_maps = preprocess_text(text, model_type, is_return_error)
filtered_text, filtered_text_number_of_words = phishing_model.filter_model_words(input_text)

filtered_text, filtered_text_number_of_words = log_io(phishing_model.filter_model_words, 'single')(input_text)
if filtered_text_number_of_words == 0:
handle_error("The model does not contain any of the input text words", is_return_error)
if filtered_text_number_of_words < min_text_length:
handle_error("The model contains fewer than %d words" % min_text_length, is_return_error)

explain_result = phishing_model.explain_model_words(
explain_result = log_io(phishing_model.explain_model_words, 'single')(
input_text,
0,
word_threshold,
Expand All @@ -171,8 +185,7 @@ def predict_single_incident_full_output(email_subject, email_body, is_return_err
explain_result['Probability'] = float(explain_result["Probability"])
predicted_prob = explain_result["Probability"]
if predicted_prob < label_threshold:
handle_error("Label probability is {:.2f} and it's below the input confidence threshold".format(
predicted_prob), is_return_error)
handle_error(f"Label probability is {predicted_prob:.2f} and it's below the input confidence threshold", is_return_error)

positive_tokens = OrderedSet(explain_result['PositiveWords'])
negative_tokens = OrderedSet(explain_result['NegativeWords'])
Expand All @@ -196,8 +209,8 @@ def predict_single_incident_full_output(email_subject, email_body, is_return_err
explain_result_hr = {}
explain_result_hr['TextTokensHighlighted'] = highlighted_text_markdown
explain_result_hr['Label'] = predicted_label
explain_result_hr['Probability'] = "%.2f" % predicted_prob
explain_result_hr['Confidence'] = "%.2f" % predicted_prob
explain_result_hr['Probability'] = f"{predicted_prob:.2f}"
explain_result_hr['Confidence'] = f"{predicted_prob:.2f}"
explain_result_hr['PositiveWords'] = ", ".join([w.lower() for w in positive_words])
explain_result_hr['NegativeWords'] = ", ".join([w.lower() for w in negative_words])
incident_context = demisto.incidents()[0]
Expand Down
Loading