Özet:

Abstract Methodologies such as Aspect Based Sentiment Classification make use of a series of texts as input in order to assess interactions pertaining to a certain object. These texts may include comments posted on social media platforms or product evaluations. Illustrations of entities include a product like a mobile phone and remarks about an entity like a restaurant. Another illustration of an entity is a restaurant. The systems look for the traits and attributes of the entity (restaurant) that are cited the most frequently, such as "service" and "food," and then attempt to determine the sentiment associated with those qualities. A number of previous technologies, for example, presented the aspect based sentiment analysis as a collection of separate subprojects, such as the aspect extraction subproblem and the sentiment assessment subtask. The purpose of this paper is to introduce a framework for an aspect-based sentiment classification as well as recommender systems. This approach will not only recognise the aspects in a highly efficient manner, but it will also be capable of performing classification tasks with a high level of accuracy using traditional machine learning techniques such as Random Forest (RF), Naive Bayes (NB), Decision Tree, Support Vector Machine (SVM), Artificial Neural Network (ANN) algorithms, and Hybrid Machine learning technique. The performance of the framework was assessed through tests on real time datasets. The framework assists tourists in finding the best venue, hotel, and restaurant in a region. It is observed from the experimental findings that the performance accuracy of proposed hybrid machine learning technique is better as compared to conventional machine learning classifier.

Anahtar Kelimeler: