Learning with N-Grams: From Massive Scales to Compressed Representations [1/3]

Channel:
Microsoft Research
Subscribers:
344,000
Published on ● Video Link: https://www.youtube.com/watch?v=ykoQ8A3pVqo


Duration: 55:58
78 views
1

N-gram models are essential in any kind of text processing; they offer simple baselines that are surprisingly competitive with more complicated state



Other Videos By Microsoft Research


2016-08-11TechFest Workshop - Theory Day - Session 1
2016-08-11Into the Mountains: The Extraordinary True Story of Survival in the Andes and its Aftermath
2016-08-11Optimizing Optimistic Concurrency Control
2016-08-11Charles River Crypto Day: On the Power of Secure Two-Party Computation
2016-08-11Pony: Co-Designing a Type System and a Runtime
2016-08-11Charles River Crypto Day: Delegating RAM Computations
2016-08-11Information-Theoretic Bounded Rationality
2016-08-11Parallel Inference and Learning with Deep Structured Distributions
2016-08-11Charles River Crypto Day: Constant-Round Interactive-Proofs for Delegating Computations
2016-08-11Stamping Out Concurrency Bugs
2016-08-11Learning with N-Grams: From Massive Scales to Compressed Representations [1/3]
2016-08-11TechFest Workshop - Theory Day - Session 4
2016-08-11General-purpose representation learning from words to sentences
2016-08-11Lab Tutorial: Multi-Objective Decision Making
2016-08-11TechFest Workshop - Theory Day - Session 2
2016-08-11The rise of the humans: at the intersection of society and technology
2016-08-11TechFest Workshop - Theory Day - Session 3
2016-08-11TechFest Workshop - Theory Day - Session 5
2016-08-11Mining Information from Microblogs during Disaster Events
2016-08-11Three Favorite Results
2016-08-11A Survey on Ring-LWE Cryptography


Tags:
microsoft research