Methods for Parsing Semi-Consistent Data

Hi Anonymous,

I think you could use Bing spell check to correct your mixing text and Text Analytic - Topic dectection to recognize the category of description as short-term solution. But it's long story cause we could not ensure it's correct 100%. Beside applying these service, you could build up and train LUIS model to understand your daily data, so it will grow up day by day to achieve what you are expecting.