The freedom of speech provides liberty to share negative, abusive, and hate speech on social media that affects mental health. Dissemination and detection of motivational and hopeful texts can motivate people in a positive direction. People with depression need a positive environment, so the detection and propagation of hope speech with natural language processing is a significant research area. We propose a novel hope speech detection framework HopeWVBL with Word2Vec embedding model (WV) for contextual analysis, POS weight for syntactical analysis, and the BiLSTM classifier (BL) to detect long dependency of text with context representation. This HopeWVBL novel approach assesses with the English language dataset that achieves 93% accuracy with a 93 F1 score and also improves 3% accuracy compared to one baseline model. The proposed model shows achievable performance against other features, classifiers, and baseline models.