没有合适的资源?快使用搜索试试~ 我知道了~
rk-ranks.ppt
1.该资源内容由用户上传,如若侵权请联系客服进行举报
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
版权申诉
0 下载量 116 浏览量
2023-06-18
21:14:14
上传
评论
收藏 1.26MB PPT 举报
温馨提示
试读
28页
rk-ranks.ppt
资源推荐
资源详情
资源评论
VLDB 2014 @Hangzhou
Reverse k-Ranks Query
Zhao Zhang, Cheqing Jin, and Qiangqiang Kang
cqjin@sei.ecnu.edu.cn
Institute for Data Science and Engineering
East China Normal University
1
VLDB 2014 @Hangzhou
2
Outline
Background
Related work
Our solutions: NA, TPA, BPA and MPA
Experiments
Conclusion
6/18/2023
VLDB 2014 @Hangzhou
3
Background
Scenario: Many applications
have a product set and a
customer set.
Each product: d-dimensional
vector (p
1
, p
2
, ..., p
d
)
Each customer: d-dimensional
preference weight (w
1
, w
2
, ..., w
d
).
For each i, w
i
>=0, and ∑w
i
=1.
∑(p
i
×w
i
) measures the preference
of a customer to a product.
Motivation: Recommend a
product to potential customers.
The distance to beach
Price
Bob
Tom
Alice
o1
o2
o3
o4
6/18/2023
VLDB 2014 @Hangzhou
top-2 hotels
4
Traditional top-k query
Consider a hotel-booking
application.
Each hotel has two
attributes, including price
and the distance to beach.
A monotonic preference
function f(.) that outputs the
score of a hotel.
Top-k query: retrieving k
hotels with the smallest
scores.
monotonic preference function
f(o) = o.distance + o.price
The distance to beach
Price
6/18/2023
VLDB 2014 @Hangzhou
5
Customized preference
6/18/2023
The distance to beach
Price
Bob
o1
o2
o3
o4
o5
o6
Each customer has different
preference.
Alice: care more about price.
f(o) = 0.1*o.distance+ 0.9*o.price
Bob: care more about the
distance to beach
f(o) = 0.9*o.distance +
0.1*o.price
Tom: in the middle
f(o) = 0.5*o.distance +
0.5*o.price
Bob and Tom like o1 very much.
But Alice prefers o2 to o1.
o7
Alice
Tom
剩余27页未读,继续阅读
资源评论
白话机器学习
- 粉丝: 8645
- 资源: 7683
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功