You are building a chatbot. You need to ensure that the bot will recognize the names of your company's products and codenames. The solution must minimize development effort. Which Azure Cognitive…

Custom text classification is used to categorize entire documents or text segments, not to extract specific entities from within the text.

Entity linking connects identified entities to a knowledge base but does not primarily focus on recognizing new, custom entity types without prior definition.

Custom Named Entity Recognition (NER) allows training a model to identify specific entities, like proprietary product names and codenames, that are unique to a company. This approach leverages machine learning to automatically extract these domain-specific terms, minimizing manual effort compared to rule-based systems.

Key phrase extraction identifies important noun phrases, but it doesn't allow for training to specifically recognize custom entity types like company product or codenames.