Text Data Labelling using Transformer based Sentence Embeddings and Text Similarity for Text Classification

This paper demonstrates that a lot of time, cost, and complexities can be saved avoided would otherwise used to label the text data for classification purposes. The AI world realizes importance labelled its use various NLP applications. Here, we have categorized close 6,000 unlabelled samples into five distinct classes. dataset was further multi-class classification. Data labelling task using transformer-based sentence embeddings applying cosine-based similarity threshold 20-30 days human efforts multiple validations with 98.4% classes correctly as per business validation. Text results obtained this fetched accuracy score F1 90%.