Difference between revisions of "Recommending What Video to Watch Next: A Multitask Ranking System (Q2678)"

From Wikibase Personal data
Jump to navigation Jump to search
(‎Created a new Item: Recommending What Video to Watch Next: A Multitask Ranking System, scholarly article from Google)
 
(‎Created claim: comment (P126): There are different types of user behaviors, such as clicking, rating, and commenting, etc. However, each one does not independently reflect true user utility. ... Our ranking system needs to be able to l...)
 
(4 intermediate revisions by the same user not shown)
Property / instance of
 +
Property / instance of: scholarly article / rank
 +
Normal rank
Property / DOI
 +
Property / DOI: 10.1145/3298689.3346997 / rank
 +
Normal rank
Property / concerns
 +
Property / concerns: YouTube / rank
 +
Normal rank
Property / date
 +
2019
Timestamp+2019-00-00T00:00:00Z
Timezone+00:00
CalendarGregorian
Precision1 year
Before0
After0
Property / date: 2019 / rank
 +
Normal rank
Property / comment
 +
There are different types of user behaviors, such as clicking, rating, and commenting, etc. However, each one does not independently reflect true user utility. ... Our ranking system needs to be able to learn and estimate multiple types of user behaviours and utilities effectively, and subsequently combines these estimations to compute a final utility score for ranking.
Property / comment: There are different types of user behaviors, such as clicking, rating, and commenting, etc. However, each one does not independently reflect true user utility. ... Our ranking system needs to be able to learn and estimate multiple types of user behaviours and utilities effectively, and subsequently combines these estimations to compute a final utility score for ranking. / rank
 +
Normal rank

Latest revision as of 18:45, 17 December 2019

scholarly article from Google
Language Label Description Also known as
English
Recommending What Video to Watch Next: A Multitask Ranking System
scholarly article from Google

    Statements

    0 references
    2019
    0 references
    There are different types of user behaviors, such as clicking, rating, and commenting, etc. However, each one does not independently reflect true user utility. ... Our ranking system needs to be able to learn and estimate multiple types of user behaviours and utilities effectively, and subsequently combines these estimations to compute a final utility score for ranking.
    0 references

    Identifiers