In the digital world, recommending the desired items to the users is still a challenging task with the unprecedented growth of the information on the Web. The recommender system plays a significant role in addressing the information overload problem in the abundance collection of the data. Recently, a contextaware recommender system has become a potential solution for providing personalized service to users. Several conventional recommendation researchers have considered contextual information in addition to the user-item-rating preferences. However, only a few recommendation research works have focused on effectively utilizing the explicit and implicit contextual information in the deep learning-based recommendations. Hence, to improve the recommendation quality, extracting the contextual user-item interaction and modeling the deep learninSg with latent context learning is essential. Thus, this work presents the DEep learning-assisted explicit and implicit COntext-based personalized Recommendation (DECOR) model to suggest the contextually desired as well as the popular items to the users. The proposed DECOR approach incorporates contextual preference extraction and deep learning-based personalized recommendation processes. Initially, it focuses on analyzing the users' explicit and implicit contexts and the items' contexts to improve the quality of the personalized recommendation. In essence, the proposed DECOR approach utilizes the explicit context of the rating, time, and location along with the implicit context from the users' reviews as the context vectors. Moreover, it enriches the individual users' rating with the items' popularity and generates the normalized rating score to facilitate the contextual and personalized recommendation. Secondly, the DECOR approach learns the contextual preferences of the users on the items with the assistance of the Long Short-Term Memory (LSTM) deep learning model. Thus, the proposed approach predicts the user preferences on the items with the knowledge of users' preference score and normalized rating score and several additional contextual factors and suggests the relevant items. Thus, the experimental results show that the DECOR approach significantly outperforms the existing recommendation model on the benchmark test dataset.