The proposed method gains domain knowledge from the vast number of documents using co-occurrence and word-embedding word vectors correlation of related data, resulting in a more coherent topic. Then, sentiment analysis is applied to reviews concerning the extracted UX aspects or dimensions. We aimed...