Multiclass Classification of Intellectual Quotes using Transformers

Paper ID: EIJTEM_2022_9_1_1-12

Author's Name: Om Verma, Kaushik Pingili , Dr. G. Arun Sampaul Thomas and S. Sathish Kumar

Volume: 9

Issue: 1

Year: 2022

Page No: 1-12

Abstract:

Many individuals enjoy reading quotes on a range of topics such as Inspiration, Motivation, Leadership, Entrepreneurship, Success, Mindset, Love, Spirituality, and more. Quotes are popular because they are concise statements that express wisdom and inspire motivation, inspiration, and happiness. Social networks have made it easier for people to connect and share their thoughts in the form of quotes, allowing individuals to inspire and help others. To better categorize these quotes based on their meaning, we have developed a transformer model. Transformers are neural network architectures that use self-attention mechanisms to recognise the relationships amid words in a sentence. This allows them to achieve state-of-the-art performance on many Natural Language Processing (NLP) tasks, including text classification. The transformer model is an innovative design that cracks sequence-to-sequence undertakings while handling long-range dependencies with ease. The use of Transformers for text classification has led to significant improvements in the accuracy and efficiency of NLP models, paving the way for more advanced and sophisticated applications of NLP in various domains. Our proposed transformer-based model for the problem of user-generated quote classification shows promising results, achieving good accuracy over traditional models for text classification problems, and has the potential to benefit both writers and readers alike.

Keywords: Transformers, Attention-mechanism, Seq2Seq Model, Text Classification.

View PDF