Rating prediction of peer-to-peer accommodation through attributes and topics from customer review

Abstract This study aims to predict customers’ behavior in classifying their reviews as high rated or low using associated attributes and topics found the review. Knowing customer reviewing action better can lead a successful strategy implementation of relevant parties related this such policy manage by keeping satisfaction high. We applied big data approach on dataset 55,377 from Airbnb listings top 10 most visited cities Indonesia (based foreign arrivals data). used The Classification Regression Tree Model, Random Forest Least Absolute Shrinkage Selection Operation Logistic Artificial Neural Network well Multi-Layer Perceptron make prediction’s classification. Those models are identify set that will increase chance review render rate different be rated. found; first, influence customers' odds classify adhere understanding Peer accommodation attributes. Second, successfully proved reviews' could classification ratings accommodation. Where for Topics, we rating yields 60.09% accuracy, slightly than (58.33%) (58.8%). However, it seems use Attributes rating, where accuracy is yielded applying with 84.79% compared only 72.35% accuracy.