Alan Mi is a Data Scientist from the Online Retail Sector. His 30-year career started with Academia, then spanning over a number of industries such as Telecom, Quantitative Investing & Trading Technologies, Airlines, and now Retail. He is specialized in Prediction, Revenue Management & Optimization, Neural Network Modeling, eCommerce Search Algorithms, and in general Data Science



Big Data for Product Search in eCommerce

Alan Mi, Data Scientist, Online Retail Sector

In this presentation Mr. Mi will dissect the product search problem in depth: the definition, a comparison against the web search problem, its innate difficulties, and search result relevance evaluation. He will then delve into semantic association representation, using Google's RankBrain as the harbinger, which calls for Big Data in Search. Next he will use a typical number crunching application - optimization - to show why Big Data is needed for small data size applications, and why companies can't throw Big Data technology at the problem brute force. Last, he will show that Big Data is needed for mining click logs, not just for the sake of coping with data size but as well to extract phantom association which feeds back into the solution for the search problem.