Rule-based annotation and an SVM (Support Vector Machine) method are the two primary parts of the implementation. According to previously established category definitions, the rule-based annotation categories words. The SVM approach uses a machine learning pipeline to anticipate words that have not yet been observed by learning from a collection of category definitions. Initially, the system reads an Excel document with category definitions. Words that fit within the categories are retrieved and stored in memory. Using the rule-based method, a dictionary is developed to map words to their appropriate categories. Additionally, Count-vectorization and a linear support vector machine model are used in a pipeline for training an SVM classifier. On the basis of the supplied word data, the SVM classifier learns to predict categories. When a paragraph is received through a Flask route, it is tokenized into sentences. The words are tokenized for each phrase, and then the categories are annotated. The rule-based technique is first used to classify terms that are already existent in the dictionary. The SVM classifier, which predicts categories based on the learnt model, processes words not covered by the rule-based method.