The tremendous growth of the Internet and the rise of social media has helped many organizations to fortunately collect voluminous data about its’ customer base. The data pertaining to customer base are the most influenceable and valuable resources for the organizations at the present time. Organizations may utilize these data to produce high revenues out of the business. Various mining techniques can be used on these data to gather valuable insights which can be used by the organization to make fascinating recommendations to their customers. These recommendations can bring in increased sales and profits to the organization. Content based recommendation is one such techniques that use item features to calculate item similarities and make recommendations. However, such techniques in a production environment use only categorical data instead of the full-fledged item-descriptions or item-review data because of their large size and increased computational resources. In this paper we provide a novel keyword-based approach for book recommendation. The text conversion technique used in the research tries to effectively reduce the text corpora while keeping the valuable information and, provides a content-based user-item rating and, fuses it with a Singular Value Decomposition (SVD) trained model to generate content lenient collaborative filtering useritem rating. The results obtained reveal that the collaborative filtering approach with keyword description has the lowest Root Mean Square Error(RMSE) score of 0.59 which is significantly lower than the RMSE score of 0.86 of collaborative filtering.