People can share their thoughts and opinions on any entities through the Internet. Normally, the attitude of the preferences of human can be predicted which are expressed in natural languages. Using sentimental mining method, the readership predictions are made on online reviews of locations. The reviews have been useful for the travelers to gain knowledge about the information of various locations and shortlist the best that is needed for them. In this paper, we categorize the locations based on the reviews and community-contributed photographs with the help of yelp and Tripadvisor datasets. In the proposed approach, opinions are filtered to eliminate unrelated ones through opinion pertinence calculation, and later grouped into a number of fused principal opinion sets. Based on the experiments conducted on large-scale datasets, the proposed approach is found to be useful for the user to make a decision. © 2020, Springer Nature Singapore Pte Ltd.