Research on how Google Autocomplete works
Budget: $50 – $0 USD
I'm looking for somebody who can help me understand how the Google autocomplete system works (so mainly: what causes a new suggestion to be shown). Ultimately, I want to understand what I can do, to cause a new suggestion to show up in autocomplete.
To this end, I would be interested in anything that would help me understand the below on a deeper level:
(1) factors involved that are used to find and rank suitable suggestions for a term in complex, large scale autocompletion, understanding the ML pipeline involved with data loading, RNN, clustering, filters, inference into a static file to show to users etc
(2) systems involved in removing suggestions due to unnatural traffic, not enough or low quality content in the index etc, not neccesarily pertaining to autocomplete
You should have know-how in that specific area but should not be bound by a work contract / NDA etc.
** You _need_ prior expert expertise with either (1) or (2). **
No definite answers are required but enough to draw up hypotheses / develop ideas that can be tested and confirmed or falsified.
No coding is required for this task but deep technical expertise.
To this end, I would be interested in anything that would help me understand the below on a deeper level:
(1) factors involved that are used to find and rank suitable suggestions for a term in complex, large scale autocompletion, understanding the ML pipeline involved with data loading, RNN, clustering, filters, inference into a static file to show to users etc
(2) systems involved in removing suggestions due to unnatural traffic, not enough or low quality content in the index etc, not neccesarily pertaining to autocomplete
You should have know-how in that specific area but should not be bound by a work contract / NDA etc.
** You _need_ prior expert expertise with either (1) or (2). **
No definite answers are required but enough to draw up hypotheses / develop ideas that can be tested and confirmed or falsified.
No coding is required for this task but deep technical expertise.