'

How Quora Uses Machine Learning to Grow the World's Knowledge

Понравилась презентация – покажи это...





Слайд 0

Machine Learning to Grow the World's Knowledge Multithreaded Data 8/18/2015 Xavier Amatriain (@xamat)


Слайд 1

Our Mission “To share and grow the world’s knowledge” • Millions of questions & answers • Millions of users • Thousands of topics • ...


Слайд 2

Our Product Teams Distribution Lookup Core Product & Quality


Слайд 3

What we care about Relevance Quality Demand


Слайд 4

Data @Quora


Слайд 5

Lots of data relations


Слайд 6

Complex network propagation effects


Слайд 7

Importance of topics & semantics


Слайд 8

Machine Learning @Quora


Слайд 9

Ranking - Answer ranking What is a good Quora answer? • truthful • reusable • provides explanation • well formatted • ...


Слайд 10

Ranking - Answer ranking How are those dimensions translated into features? • Features that relate to the text quality itself • Interaction features (upvotes/downvotes, clicks, comments…) • User features (e.g. expertise in topic)


Слайд 11

Ranking - Feed • Personalized learning-to-rank approach • Goal: Present most interesting stories for a user at a given time • Interesting = topical relevance + social relevance + timeliness • Stories = questions + answers


Слайд 12

Ranking - Feed • Features • Quality of question/answer • Topics the user is interested on/ knows about • Users the user is following • What is trending/popular • … • Different temporal windows • Multi-stage solution with different “streams”


Слайд 13

Recommendations - Topics Goal: Recommend new topics for the user to follow • Based on • Other topics followed • Users followed • User interactions • Topic-related features • ...


Слайд 14

Recommendations - Users Goal: Recommend new users to follow • Based on: • Other users followed • Topics followed • User interactions • User-related features • ...


Слайд 15

Related Questions • Given interest in question A (source) what other questions will be interesting? • Not only about similarity, but also “interestingness” • Features such as: • Textual • Co-visit • Topics • … • Important for logged-out use case


Слайд 16

Duplicate Questions • Important issue for Quora • Want to make sure we don’t disperse knowledge to the same question • Solution: binary classifier trained with labelled data • Features • Textual vector space models • Usage-based features • ...


Слайд 17

User Trust/Expertise Inference Goal: Infer user’s trustworthiness in relation to a given topic • We take into account: • Answers written on topic • Upvotes/downvotes received • Endorsements • ... • Trust/expertise propagates through the network • Must be taken into account by other algorithms


Слайд 18

Trending Topics Goal: Highlight current events that are interesting for the user • We take into account: • Global “Trendiness” • Social “Trendiness” • User’s interest • ... • Trending topics are a great discovery mechanism


Слайд 19

Spam Detection/Moderation • Very important for Quora to keep quality of content • Pure manual approaches do not scale • Hard to get algorithms 100% right • ML algorithms detect content/user issues • Output of the algorithms feed manually curated moderation queues


Слайд 20

Content Creation Prediction • Quora’s algorithms not only optimize for probability of reading • Important to predict probability of a user answering a question • Parts of our system completely rely on that prediction • E.g. A2A (ask to answer) suggestions


Слайд 21

Models ● Logistic Regression ● Elastic Nets ● Gradient Boosted Decision Trees ● Random Forests ● Neural Networks ● LambdaMART ● Matrix Factorization ● LDA ● ...


Слайд 22

Conclusions • At Quora we have not only Big, but also “rich” data • Our algorithms need to understand and optimize complex aspects such as quality, interestingness, or user expertise • We believe ML will be one of the keys to our success • We have many interesting problems, and many unsolved challenges


Слайд 23


Слайд 24


×

HTML:





Ссылка: