This repository contains the code and resources for my internship project, where I leveraged the Roberta model for sentiment analysis. During my internship, I explored and implemented sentiment analysis tasks using state-of-the-art natural language processing techniques.
The dataset was provided by internship company. It contains 2 columns: an ID column and a column that shows the review of the customer. Here is a preview of the dataset.
The success of sentiment analysis tasks heavily relies on the quality of the underlying language model used for processing text data. In this project, I employed the RobeRata model as the cornerstone of our sentiment analysis pipeline.
The Roberta model's adaptability and accuracy in processing Romanian text make it a valuable asset for sentiment analysis tasks in the Romanian language. By leveraging its capabilities, we aim to provide precise sentiment analysis results for various applications within our project.
The results of this analysis are presented in this powerpoint