1. Fake reviews and inflated ratings is a big concern for e-commerce platforms. A fake review is a review written by someone who has not actually used the product or given manipulated review to get refund or the offered incentive. (a) Suppose you aim to build a model to identify fake reviews. Unfortunately, we do not have any existing labelled dataset for this problem. We also lack sufficient human annotators to label big enough dataset for supervised learning. • Design a partially/fully automated solution to create labelled dataset. • How would you ensure the quality of the data? • Discuss the pros and cons of the proposed solution. (b) Design an approach to flag fake reviews. Clearly demonstrate all the modules using a diagram.
1. Fake reviews and inflated ratings is a big concern for e-commerce platforms. A fake review is a review written by someone who has not actually used the product or given manipulated review to get refund or the offered incentive.
(a) Suppose you aim to build a model to identify fake reviews. Unfortunately, we do not have any existing labelled dataset for this problem. We also lack sufficient human annotators to label big enough dataset for supervised learning.
• Design a partially/fully automated solution to create labelled dataset.
• How would you ensure the quality of the data? • Discuss the pros and cons of the proposed solution.
(b) Design an approach to flag fake reviews. Clearly demonstrate all the modules using a diagram.
Step by step
Solved in 5 steps