DBTB INT Vita Markman

I am Vita Markman I'm a staff software engineer at LinkedIn meeting so many different people from the profession of data science and text analysis and natural language processing and going to their talks because there is so much of it and it is so sparse than the one hand and there's but it's the quantity is great but the noise is all so overwhelming and to find the right meaning in the data so that serve a business purpose is really challenging that machine learning is not magic learning and you need to have quality data if you have supervised student algorithm that the data has to be clean and consistent you have to study the data you cannot assume that data is given to you either by crowdsourcing or by some magical elves and the data is of good quality you the practitioner have to assure that your data is good by doing the proper vetting and if it's not good you have to be the one who fixes it you can't assume that it just happens or that the algorithm will somehow magically take care of it it will never happen you