Help with understanding an Algorithm

I'm trying to code the Algorithm discussed in the paper - (see here http://tinyurl.com/nzdf6rq ). The problem is I am not sure if I understand the problem enough to be able to correctly write the code.

This is what I understand from the problem so far. I am hoping someone can help fill in blanks.

To determine a trending activity:

- Take a recent activity (observation s)
- Compare s to the example patterns r (from past topics)
- For each example, take a vote V(r,s) to determine if a topic is trending or not.


Assuming my understanding is correct, what other information do we require in the example beside the name of the topic that we are tracking?

what variables are we to use to compare s with r?

What is the purpose of scaling parameter y (greek symbol)?

It would be a real help if someone can break the algorithm down into a series of steps that is easy to follow.

Quote:

---

Originally Posted by Hallow

It would be a real help if someone can break the algorithm down into a series of steps that is easy to follow.

---

I think your best bet is to contact the author actually. If you show a keen interest in someone's work they're often very willing to discuss it with you.

Aye, I strongly agree with nuzzle. People love it when people take an interest in their work...!