What is UPC Hunter?

[Tripredacus]

I found the super secret Shmax UPC Hunter! 

While it is neat, I can search for UPC codes, it seems to only link to products on Amazon.com. Both images and links... and it has items that are NOT in the Shmax database.

Anyways, since I couldn't find a link to it from anywhere on the site... I won't post the link.

[shmax]

I found the super secret Shmax UPC Hunter! 

While it is neat, I can search for UPC codes, it seems to only link to products on Amazon.com. Both images and links... and it has items that are NOT in the Shmax database.

Anyways, since I couldn't find a link to it from anywhere on the site... I won't post the link.

It's a tool I cranked out one night for digging UPCs out of Amazon's database (they no longer seem to expose them on their own site, for some reason) via their API. It's just an experiment that I never followed up on, and I don't think anyone around here ever really got in the habit of using it. I'll probably just switch it off if no one objects.

Oh, and you've always been able to search shmax.com by UPC. Just type 'em into the search box at the top of the page. If you have a bar code reader, it's a good way to quickly enter in your collection.

[Antron007]

 Oh neat. Could you post a link Max? Let me drive  it for a couple days before you put it in the garage? I'd like to check that out.

[shmax]

Oh neat. Could you post a link Max? Let me drive  it for a couple days before you put it in the garage? I'd like to check that out.

http://www.shmax.com/upc_hunter

but I think I limited it to mods only. I'll switch it to logged-in users only when I get home and you can play around.

[Antron007]

Thanks Max :-) :-) I just want to take it out for a spin or two just to see what I find. As always, you rock!!

[Antron007]

That's a nifty little tool. You ever think about making a script that runs the UPC search and compares it to the Shmax DB? Maybe add the comparison part to the existing UPCs page and have it only display missing entries, or flag them. Once again, just thinking out loud for you.

[shmax]

That's a nifty little tool. You ever think about making a script that runs the UPC search and compares it to the Shmax DB? Maybe add the comparison part to the existing UPCs page and have it only display missing entries, or flag them. Once again, just thinking out loud for you.

I could maybe resurrect the ebay auction matching code to try to pair up amazon records with shmax records, but we'd still need a human to validate each match. Since a human would need to be involved anyway, my solution was to just make a tool for humans to use. The humans never did, but I can't really blame them; I've used the tool myself enough to know that it can be difficult to find the more obscure toys on Amazon; all the low-hanging fruit has already been UPC'ed in our db.

[Antron007]

Similar but side note, I notice on the contribute page, it only shows things I have. Is there a fake user with the entire DB added to it so we could see a complete list of missing info?

And more on topic, is there a way to generate a list of all of the upc's in the DB? if someone had that they could collect the upcs from the hunter and compare them. That should find missing ones that will have to be checked from there but it would cut out physically matching records for existing upcs in the DB.

[shmax]

Similar but side note, I notice on the contribute page, it only shows things I have. Is there a fake user with the entire DB added to it so we could see a complete list of missing info?

Ah, we used to have a link for seeing everything. Until I get around to restoring it, you can use this:
http://www.shmax.com/missing_info/1?info_filter=4095. It's actually a bit chuggy--maybe it's best not too many people know about it for now.

[Antron007]

If you wanna pull it and PM it to me that's be cool. Unless trip wants in on this at all. It's sort of his fault. He started it. ;-p

I think you're ebay code may be a little better than you think. I've tried manually searching ebay for upcs with little luck. It really seems to suck for them. At least for me. I think applying the code to Amazon may work out pretty good.

Have you ever coded the UPC hunter for ebay instead of Amazon?

[shmax]

If you wanna pull it and PM it to me that's be cool. Unless trip wants in on this at all. It's sort of his fault. He started it. ;-p

I think you're ebay code may be a little better than you think. I've tried manually searching ebay for upcs with little luck. It really seems to suck for them. At least for me. I think applying the code to Amazon may work out pretty good.

Have you ever coded the UPC hunter for ebay instead of Amazon?

I think there may be a little bit of a misunderstanding. The purpose of the UPC hunter tool is to expose the UPCs of items in Amazon's database. They do have the numbers for a lot of things, but they don't display them anywhere on the site (or at least they didn't back when I wrote the tool). They are, however, available in the metadata you can get back from talking directly to their database through their API service (I have a developer key from my participation in their affiliate program). So, the idea is that you can use my tool to query their database and dig around through the results looking for UPCs that we need for shmax.

I only mentioned eBay because I wrote an algorithm that can match auction titles with items in the shmax database, and it's feasible that I could use the same technology to try to match up amazon product names with shmax items for the purposes of saving some time. But whereas we can allow a little room for error with auction matching for the purposes of generating graphs, it would have to be spot-on for gathering UPCs, so a human would have to be involved in the process. I could write a script that sort of generates a report of best guesses for a human to look over and confirm/reject, but I'm busy with other stuff now, and the users are already doing a great job of supplying that info.

One doesn't look up UPCs on eBay, because eBay doesn't manage their own archive of specific toys (yet, knock on wood--they do have a database of CDs, DVDs, and video games), nor do sellers generally enter them.

[Antron007]

I've found that on ebay, even when I enter the UPC, if you preview or make edits to your auction it deletes the UPC info . So even when people think the UPC is there at one point, if they edit and don't pay attention, it will be gone. I think that might have a little to do with it.

I have more on this but I gotta jet but in short I think we are on the same page, just different paragraphs. ;-)

[Antron007]

Ok I got a sec. I know there will have to be human interaction but what I'm suggesting is simply taking all upcs in shdb (X) and comparing it to the upcs generated by the upc hunter (Y).

Y-X= UPCs not in shmax. That's the theory anyway. Then one could take that list, (While possibly not perfect) and use it to fill in missing info in X. So at some point Y-X should = 0 because all the Y info previously not in X would now be there.

That's what I'm trying to suggest as simply as possible. It's something anal enough that I may spend at least a few minutes a day inputting upcs into the db.  I suppose I could do it all manually now that I have a link to the missing shdb info (Thanks again for that. I'm going to try my best to take advantage of it.) I can copy all the upcs into a speadsheet and do the same for the amazon data then compare them.

I just think being able to compare the two datasets would help us quickly fill in some holes. I'll try to whip something up in the next week or so. I should be able to take the info and compare them with a script of some sorts. It's been a while but I used to take google searches and reformat them. Something about capturing the data stream and breaking it into it arrays then re-displaying the elements as I saw fit.

In theory I should be able to catch both streams, sort them by UPC then compare them then display a list that amazon has that shdb doesn't.  We'll have to verify that list from there but I think it would eliminate alot of the sorting and the need to look at products we already have the upcs for.

I hope I'm not making this complicated because it really sounds simple in my head until I start typing it all out.

Any thoughts from there?

[shmax]

Ok I got a sec. I know there will have to be human interaction but what I'm suggesting is simply taking all upcs in shdb (X) and comparing it to the upcs generated by the upc hunter (Y).

Y-X= UPCs not in shmax. That's the theory anyway. Then one could take that list, (While possibly not perfect) and use it to fill in missing info in X. So at some point Y-X should = 0 because all the Y info previously not in X would now be there.

That's what I'm trying to suggest as simply as possible. It's something anal enough that I may spend at least a few minutes a day inputting upcs into the db.  I suppose I could do it all manually now that I have a link to the missing shdb info (Thanks again for that. I'm going to try my best to take advantage of it.) I can copy all the upcs into a speadsheet and do the same for the amazon data then compare them.

I just think being able to compare the two datasets would help us quickly fill in some holes. I'll try to whip something up in the next week or so. I should be able to take the info and compare them with a script of some sorts. It's been a while but I used to take google searches and reformat them. Something about capturing the data stream and breaking it into it arrays then re-displaying the elements as I saw fit.

In theory I should be able to catch both streams, sort them by UPC then compare them then display a list that amazon has that shdb doesn't.  We'll have to verify that list from there but I think it would eliminate alot of the sorting and the need to look at products we already have the upcs for.

I hope I'm not making this complicated because it really sounds simple in my head until I start typing it all out.

Any thoughts from there?

Apologies for the absurdly late reply, but hmm, that's actually a very interesting idea. I may give it a shot later this evening...