Client

Rowan University

Year

2017

Role

Researcher

Technologies Used

Python, TensorFlow, Scikit-Learn, Neural Networks, Word2Vec

Deceptive opinionated spam are synthesized reviews designed to promote products or demote competitors. It is filled with fictitious opinions, deliberately written to sound authentic. Detecting this spam is deemed a classification problem, supervised learning builds a classifier which then predicts if it is a deceptive review or a truthful review.

Received 90% accuracy using our method.

Final presentation from T K.