Highly matched semantic answers to natural language queries.
Many organizations have an existing set of frequently asked questions that human support agents are trained on. For example, if a user encounters a problem on his mobile phone, then the manufacturer has established steps the user should follow to correct the problem. Using NLP, IDOL identifies the FAQ that best answers a query.
In many cases, the information requested is simply not present in either an FAQ data set or a structured database, so an extended approach is required. IDOL passage extraction looks through the corpus of data to find segments of documents that best answer the query directly in short sentences with relevant answers.
