Which of the following controls helps mitigate the risk of competitors poisoning data utilized by a machine learning (ML) model performing sentiment analysis of product reviews?

Question

Options

• APeer reviewing code that acquires product reviews from social media posts
• BHiring a marketing firm to text links to customers requesting product reviews for monetary
• CAugmenting the unbalanced product review data set with the use of oversampling by the model
• DRequiring customers to authenticate access to their accounts prior to writing product reviews

• A
  13% (7)
• B
  5% (3)
• C
  4% (2)
• D
  79% (44)

Explanation

Data poisoning attacks involve injecting malicious or fabricated data into a model's training set to skew its outputs. Requiring customers to authenticate before submitting reviews ensures that only verified, real customers can contribute data, blocking competitors and bots from flooding the system with fake reviews that could corrupt the sentiment model. Option A (code peer review) is a software quality control and does not prevent unauthorized data injection. Option B (paying for reviews) introduces ethical and legal risks and its own form of bias. Option C (oversampling) addresses class imbalance in datasets, not adversarial data injection.

No community discussion yet for this question.