Patient complaints are crucial for medical diagnosis and treatment planning, yet they are often recorded in unstructured text, making analysis challenging. This study applies statistical information extraction to classify patient complaints into structured components using the Support Vector Machine (SVM) algorithm. The dataset consists of transcribed complaints from clinical consultations and an online medical platform. Preprocessing involves eleven steps, including case folding, punctuation removal, number-to-text conversion, abbreviation normalization, spell correction, stopword removal, tokenization, n-gram processing, and array indexing. Feature extraction is performed using Term Frequency-Inverse Document Frequency (TF-IDF), followed by classification using SVM with Stochastic Gradient Descent (SGD). The system’s accuracy is 56.75%, with a precision of 41%, a recall of 53%, and an F1 score of 42%. Expanding the dictionary increases accuracy to 86%, highlighting the importance of a comprehensive lexical database. Despite these enhancements, limitations persist, such as the need for a larger dictionary and a dynamic vocabulary editing feature. Future research should explore deep learning models to enhance classification accuracy and validate the system on a larger dataset. This study contributes to the development of automated systems for structuring patient complaints, improving medical documentation, and clinical decision-making.