We recently translated data to train an AI chatbot. The customer was an advertising agency, which in turn was hired by the ticketing service. In total, the project implemented 4,500 question-and-answer scenarios for site users. For each scenario, our linguists had to come up with 15-20 ways to ask the same question, using different tags and keywords and without repeating sentences. The initial information was the answer to these possible questions.
Chatbots using machine learning need to be provided with a lot of training data. You need to write hundreds or even thousands of statements that reflect what your target user might ask the chatbot and how they might ask it.
In fact, many of the world`s largest translation agencies are already promoting machine learning services. The artificial intelligence industry is only set to grow, and the demand for data to train chatbots is another example of how technology is creating new opportunities for translators.
