Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

After looking at a number of these algorithms, it seems like you really need to take each situation into consideration. I like that Google example as it is easy and scalable as you said. But for more complicated situations, you can do batch updates using Hadoop/MapReduce, assuming you don't have popular items that need to be calculated real-time.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: