Named Entity Recognition Using BERT Model for Kannada Language

Named Entity Recognition (NER) is a vital technique in Natural Language Processing (NLP) used to discern and classify words in a text into corresponding predefined categories. These categories often inlude entities such as names of people, organizations, locations, dates, and more. These are termed as Named Entities. The goal of NER is to extract and label these named entities automatically from unstructured text data. It has many applications, especially in information extraction, machine translation, text-to-speech synthesis, natural language understanding, question answering, text summarization, etc. This paper presents an approach based on Bidirectional Encoder Representations from Transformers (BERT) for identifying and recognizing named entities in Kannada, an Indian language.