Stanford Seminar - The Case for Learned Index Structures

Transcript

Read and summarize the transcript of this video on Glasp Reader (beta).

Summary & Key Takeaways

The summary of this video is not yet available 😢 We summarize videos one by one, so it might take a while. If you want to get the summary now, please summarize the video with YouTube Summary extension or try Glasp Reader.

About the Video

EE380: Computer Systems Colloquium Seminar

The Case for Learned Index Structures

Speaker: Alex Beutel and Ed Chi, Google

Indexes are models: a B-Tree-Index can be seen as a model to map a key to the position of a record within a sorted array, a Hash-Index as a model to map a key to a position of a record within an unsorted array, and a BitMap-Index as a model to indicate if a data record exists or not. In this talk, we take this premise and explain how existing database index structures can be replaced with other types of models, which we term learned indexes. The key idea is that a model can learn the sort order or structure of indexed data and use this signal to effectively predict the position or existence of records. We offer theoretical analysis under which conditions learned in...