Portfolio
Sentiment Analysis: Lexical Based Models vs Deep Learning Models
-
Client Personal
-
Category AI Engineering
-
Date July, 2023
In this university group project, under my leadership, we embarked on an in-depth exploration of sentiment analysis techniques, aiming to identify the most effective method for accurately determining the sentiments expressed in text data. Our research focused on comparing the performance of Lexical Based models against that of advanced Deep Learning models. The primary objective was to assess which approach could deliver superior accuracy in analyzing sentiments.
Our methodology included:
- Comprehensive evaluation of Lexical Based models, which rely on predefined dictionaries of words associated with positive or negative sentiments. These models were tested for their ability to accurately categorize sentiments based on the presence of these words within the text.
- Implementation and testing of various Deep Learning models, including Convolutional Neural Networks (CNNs) and Long Short-Term Memory networks (LSTMs), known for their capability to understand and interpret the context of words in sentences more effectively than traditional methods.
- Rigorous comparison of the results, which revealed a stark contrast in performance: Deep Learning models significantly outperformed Lexical Based models, achieving an astounding 99% accuracy rate compared to the 64% accuracy rate of Lexical Based models.
This project highlighted the profound impact of Deep Learning techniques in the field of sentiment analysis, showcasing their superior ability to capture and interpret the nuances of human language. As a team leader, I steered our group through the complexities of model selection, implementation, and evaluation, culminating in a successful demonstration of the potential for Deep Learning models to revolutionize sentiment analysis tasks. This research not only provided valuable insights into the comparative effectiveness of different sentiment analysis methodologies but also underscored the advancements in machine learning and its applications in understanding human emotions through text.