In recent years, the cold start problem and accuracy has posed significant challenges in recommender systems, particularly for new users with limited interaction history. The contemporary age of recommendations has been substantially improved by the application of sentiment analysis. The incorporation of trust and reputation-based recommendation has the potential of improving the accuracy and user acceptance towards the presented recommendation. This paper addresses the issue of new user cold start problem by integrating sentiment analysis with trust and reputation metrics that enhance recommendation accuracy and handles the cold start problem. By incorporating trust with sentiment score improves the user acceptance towards the presented recommendations. The proposed approach leverages sentiment scores from user reviews, obtained through SVM and logistic regression, and combines them with trust and reputation measures to provide a more accurate recommendation. Collaborative filtering techniques are used to create user similarity matrices based on these sentiment scores. By removing users with negative similarity scores and combining trust and sentiment-enhanced scores, the recommendation system’s performance has been targeted. The results indicate that incorporation of sentiment analysis and trust metrics mitigates the new user cold start problem, with more accurate and personalized recommendations for new users.