Table 1 Summary of the reviewed papers in which no validation procedure has been applied to examine the relationship between numeric ratings and texts reviews
From: Text-rating review discrepancy (TRRD): an integrative review and implications for research
| Â | Reviewed Paper | Used Model | Domain | (Manually labelling) Reviews were labelled by Experts? | (Automatic labelling) Reviews were labelled by Sentiment Lexicons? |
|---|---|---|---|---|---|
Survey Papers | Shoham et al. [51] | Survey Study | – | – | – |
Sentiment Analysis using Machine Learning | Kim et al. [28] | Binary classification | Movies | No | No |
Rui et al. [46] | Multiclass classification | Movies | No | Yes | |
Hamouda et al. [22] | Binary classification | Amazon from different domains (books, cameras, mp3s, etc.) | No | No | |
Pang and Lee [39] | Multiclass classification Regression | Movies | Yes | No | |
Pang et al. [40] | Binary classification | Movies | No | No | |
Sentiment Analysis using Different Techniques | Tang et al. [57] | Neural Network + user-word composition vector model (UWCVM) | Restaurants and movies | No | No |
Basiri et al. [3] | Lexicon-based model | Restaurants and hotels | Partially | Yes |