Predictive model using youtube and/or googletrends.

A forum for general discussion of the Python programming language.

Predictive model using youtube and/or googletrends.

Postby BDD » Fri Apr 04, 2014 12:53 pm


I am new to these forums and registered because I have a few questions regarding a University project I am currently working on.

We are aiming to build a predictive model that basically does the following:

-Gets youtube tag data from the top viewed videos in specific youtube categories over a period of a week.
-List these tags according to popularity and for example save the top 100/500/whatever.
-Crossreference these top tags with google trends performance of these tags.
-End up with a list of most talked about products

I was wondering if something like this is even possible in python and if people could give pointers as to how I would achieve building this. I have been toying around with dozens of scripts, api's and other prewritten code as we are quite a novice team but I have not even come close to something like the model as we envision it.
Last edited by stranac on Fri Apr 04, 2014 1:05 pm, edited 1 time in total.
Reason: First post lock.
Posts: 1
Joined: Fri Apr 04, 2014 12:49 pm

Re: Predictive model using youtube and/or googletrends.

Postby micseydel » Sat Apr 05, 2014 12:10 am

Vague questions like this can't get better than vague answers, I'm sorry to say. The basic answer is "yes". Of course we'll try harder than that...

Are you a programmer? If not, are you technical? Are you considering any other language(s) to do this same thing? What is your time frame on all this? If the computations are very intensive, do you have access to hardware that could allow you to parallelize the effort? (This scales better than writing parts of your code in C, although writing parts in C might also be a viable solution to slowness. PyPy could also potentially be helpful.) You may want to consider seeking a computer science student at your university, or perhaps someone else who is used to using programming to do math (statistics students could oftentimes use R); from my experience it's difficult to find anyone motivated enough for hard problems like this, but you could try.

When people come with questions like this, it's must like going up to someone in NASA and saying, "Hey, I'd like to build a rocket ship. Where do I start?" And the first step, of course, is a ton of research and learn... everything. Like, 4+ years is what you'd expect to be able to learn it from someone else who already knows.
When you have a large task that you're aiming to get done, it can be very hard to learn the tools you need to get that task done because you're so confused on the big picture. You'll probably learn not enough, try to do your task, and in some way fail or screw up. I don't mean this as a criticism, or even as something to even avoid, it just seems to be part of a hard reality -- if you don't know how to do this, it's probably going to be very, very, very hard. For any of us would probably just be very hard (though perhaps more information would reveal that it is just hard, or even easy; it's difficult to tell). Once you kind of get started, you'll be able to ask more specific questions, and we'll be able to give more specific answers.

You'll also probably benefit from reading this. We really would like to help you here, so if I've accidentally offended you, I welcome any criticism toward me so that I can learn to better help you and others with questions like this.
Due to the reasons discussed here we will be moving to on October 1, 2016.

This forum will be locked down and no one will be able to post/edit/create threads, etc. here from thereafter. Please create an account at the new site to continue discussion.
User avatar
Posts: 3000
Joined: Tue Feb 12, 2013 2:18 am
Location: Mountain View, CA

Return to General Discussions

Who is online

Users browsing this forum: No registered users and 4 guests