Add SVM-based sentiment analysis

[RachelMMMPPP]

This pull request introduces a sentiment analysis feature using Support Vector Machines (SVM). I apologize if this contribution is not appropriate; it was initially created for my machine learning class assignment. Please feel free to disregard this request if it does not fit with the repository's goals. I sincerely apologize for any inconvenience caused.
The new implementation includes the following changes:

1. Data Handling: Utilizes the provided dataset (tweets.csv) for training the SVM model to predict sentiments (positive, negative, or neutral).
2. Preprocessing: Handles missing values in the dataset by replacing them to ensure clean input data for model training.
3. Vectorization: Implements text vectorization using TfidfVectorizer to convert textual data into numerical format suitable for SVM classification.
4. Model Training: Trains an SVM classifier using the preprocessed data, enabling the prediction of sentiment based on textual input.
5. Testing: Added unit tests to verify the correctness of the sentiment prediction, ensuring the model accurately classifies input text into one of three sentiment categories: positive, negative, or neutral.
6. Documentation: Included comments and docstrings to explain the code's functionality and usage.