Designing a Query Auto Completion System

This was a part of our tutorial on "Deep Learning for AutoSuggest" tutorial at IJCAI 2022. Part 1: https://youtu.be/eyJSqrxwwGU Part 2: https://youtu.be/7fGnZaifLoE Part 3: https://youtu.be/dO_AJ2Wn0Z0 Part 4: https://youtu.be/nboTXqZTG_I Part 5: https://youtu.be/bNpiTFzJkxs Query Auto Completion (QAC) aims to help users reach their search intent faster and is a gateway to search for users. Everyday, Billions of keystrokes across 100s of languages are served by Bing Autosuggest in less than 100 ms. The expected suggestions could differ depending on user demography, previous search queries and current trends. In general, the suggestions in the AutoSuggest block are expected to be relevant, personalized, fresh, diverse and need to be guarded against being defective, hateful, adult or offensive in any way. In this video, I broadly talk about important components in designing an effective query auto completion (autosuggest) system. Here is the overall agenda 00:00 Ranking suggestions 04:59 Most popular completion 09:57 Time sensitive suggestions 13:00 Location sensitive suggestions 15:58 Personalization 17:19 Ghosting, Session co-occurrences 20:21 Online spell correction, Defect handling 23:28 Non-prefix matches, Generating suggestions 26:32 Mobile QAC, Enterprise QAC For more details, please look at https://drive.google.com/file/d/1h0nxun9nLOKs47p8zfUupVK9pmhiT_24/view Tutorial: Manish Gupta, Puneet Agrawal. Deep Learning Methods for Query Auto Completion. The 31st International Joint Conference on Artificial Intelligence and the 25th European Conference on Artificial Intelligence. July 23-29, 2022. Vienna, Austria.