The goal of most sentiment tasks is to identify the overall sentiment polarity of the documents in question, i.e. is the sentiment of the document positive or negative? For our case, we will enter sentence in Azerbaijani and model will predict its sentiment. In order to perform this sentiment task, we use a mixture of baseline machine learning models and deep learning models to learn and predict the sentiment of binary reviews.
I have implemented Multi Layer Perceptron model to learn and predict the sentiment of sentences
This poses a supervised learning task. For this purpose, we will use bag-of-words features with machine learning algorithms which is very simply and efficient, while having the ability to achieve very high accuracy. We will develop a neural bag-of-words model, which collects high-level structural and semantic meaning of the words.
The project performs sentiment classification via neural bag-of-words approach. After preparing data and building model, I have deployed this project using Flask.