B*******1
发帖数: 2454 | 1 Amazon.com has the feature that if users search for particular CD or books
etc then it also tells what other related books/CDs users bought.
Example, if a user search for the book "Introduction to Algorithms" then it
also shows users who bought this book also bought "Robert Sedgewick -
Alogrithms in C++" or "Fundamentals in Data Structures".
Assuming everything is maintained in memory & file system what algorithm
will give this set of related intersection/connection ? |
d*******d
发帖数: 2050 | 2 我觉得这是个图论结构题,问的是你要怎么建立这个图.
algorithm好像用不上什么特别的.
it
【在 B*******1 的大作中提到】
: Amazon.com has the feature that if users search for particular CD or books
: etc then it also tells what other related books/CDs users bought.
: Example, if a user search for the book "Introduction to Algorithms" then it
: also shows users who bought this book also bought "Robert Sedgewick -
: Alogrithms in C++" or "Fundamentals in Data Structures".
: Assuming everything is maintained in memory & file system what algorithm
: will give this set of related intersection/connection ?
|
B*******1
发帖数: 2454 | 3 我也觉得是，
譬如A书，用户甲买了A+B，用户乙买了A+C
amazon怎么提示B还是C啊，是不是A和B之间的weight根据多少用户买了A+B来定，然后
给提示的时候就根据所有从A出来的的edge给出weight最高那个node啊？ |
g**********y
发帖数: 14569 | 4 If A buy book b, find all users buy book b. Count these users' order, find
top two books to recommend.
In real, I believe they have better logic, like restrict books to same
category of book b.
★ 发自iPhone App: ChineseWeb - 中文网站浏览器
【在 B*******1 的大作中提到】
: 我也觉得是，
: 譬如A书，用户甲买了A+B，用户乙买了A+C
: amazon怎么提示B还是C啊，是不是A和B之间的weight根据多少用户买了A+B来定，然后
: 给提示的时候就根据所有从A出来的的edge给出weight最高那个node啊？
|
