Sentiment analysis is one of the most common tasks in text analysis. Often companies want to know what a customer’s sentiment towards a business process, product, experience, or aspect of their company. Customers tend to make comments about companies they interact with in a variety of ways including through social media, emails, surveys, etc… The problem is, how can a company determine the sentiment towards aspects of its business efficiently in an automated way using only social media posts?
Solution:
Develop a supervised classification model to predict the sentiment of a sentence of text.
Methods:
Deep neural network embeddings, transfer learning using pre-trained models and data set augmentation
Frameworks and Platforms:
Python, scikit-learn, spacy, etc…
Outcomes:
I developed a high accuracy (over 95%) pipeline for classifying the sentiment of a social media posts.
