Business Challenge: For a business website containing humongous frequently asked question data, it at times get difficult to cater to user queries. This results in slower customer response service as well as it may adversely affect the customers on the website.
Based on the question posed by the customer, the semantically related query would be generated and the previous FAQ closest to the customer query would be shown.
By generating sentence embeddings of the previously frequently asked questions, data cosine distance would be calculated between the two sentence embeddings.
For the newly asked questions by the customers, the closest frequently asked questions from the previous data set would be presented to the customer.
This would save customers time and provide more efficient FAQs service, resulting in improved customer service.