Suicide is a critical and serious public health issue. Prevention and detection of suicidal thoughts are essential to save an individual. Suicide attempt prediction aims to recognize the user is in the suicidal intention based on their information shared in the social media. Artificial intelligence is the emerging technology for investigating large datasets to enhance risk detection. Deep learning is the recent generation of AI technologies. Deep learning is a suitable learning model that automatically detects suicidal behavior and thoughts by extracting the features from the online social contents of an individual.
   The deep learning model is solely inadequate for predicting the suicidal thoughts of an individual. An additional lexicon knowledge is required for suicidal thoughts prevention which helps in better understanding of suicidal thoughts. The lexicon model analyses the physiological behavior of an individual using its preserved knowledge. Artificial intelligence and lexicon-based suicide attempt prevention helps in early suicidal intention understanding and boost predictive performance.