Applying NLP Techniques on Search Queries
Understanding user-generated inputs in information retrieval systems is very challenging due to minimal context and few structural clues. Lyst is a marketplace operating in Fashion with the goal to provide the best user experience. As Lyst operates in a restricted domain, Named-Entity-Recognition algorithms can be useful to understand search queries semantically, and retrieve all the relevant products the users are looking for. Additionally, NER algorithms can provide feedback to users to disambiguate their search queries. For example, the search query "red valentino dresses" is vague. Does the user want dresses by Red Valentino or they want red dresses by Valentino?
Pavlos is a Senior Data Scientist at Lyst (a personalized fashion marketplace), where he focuses on Search Retrieval, Learning2Rank and applying NLP techniques on Search. Pavlos has an MSc in Computer Science from Imperial College London. Previously, Pavlos was a Data Scientist at Workable designing and building Identity Resolution algorithms and Recommender Systems. Before that, Pavlos worked as a Software Engineer at Expedia building an Automation/Bidding Platform for bidding in AdWords.