Take the 2-minute tour ×
Webmasters Stack Exchange is a question and answer site for pro webmasters. It's 100% free, no registration required.

I am requested to analyze the traffic log of a site. In particular I have to identify the crawlers starting from the collected user agent values.

I know there are 'trap' links that you can use to distinguish the crawlers from the human beings. Now I would only analyze the user agent values.

Now the question. Is there a public catalogue or a library of web crawlers?

Edit

Here is the second question. There are also a lot of empty user-agent in my traffic records. Is an empty user-agent header related to a crawler or to an authomatic process?

share|improve this question

3 Answers 3

The first link on a Google search is probably what you are looking for - http://www.user-agents.org/

share|improve this answer
    
Good starting point. Who do maintain the db? Up to when the data are updated? –  ADC Aug 7 '12 at 8:25
    
@ADC There are timestamps in at least two locations on that page, although they're not terribly recent. But it's one resource to have around. The forum(feedback & additions) seems to still have some people making submissions, even if they may not be getting published to the site. –  Su' Aug 7 '12 at 9:50

It's highly unlikely you're going to find some completely universal list of UserAgents, in part because they can just be made up. Before even getting to that, though, it'd be a ridiculous amount of work. You just need to compile a few resources and then do some further searching for anything else you don't recognize. (Surprisingly, I can't find a Wikipedia "List of…" article for this.)

  • Here's a massive list of nothing but iOS UA strings. If you look at how fast some of those get changed in the date column and take into account the last update to the document was 10 weeks ago, it's quite possibly already missing something.
  • UserAgentString.com seems more recently-maintained than user-agents.org. Every one of those product names leads to a separate page with its own sometimes-huge list.
  • user-agent-string.info has a lot of non-browser ones that appear to be missing from the previous, so might also be good to have around.
share|improve this answer
    
The Apple devices do share a common string being Apple-iPhone and Apple-iPod etc. I don't think looking for all bots in raw access logs would work either especially since there's bound to be a few malicious bots using standard browser user agents. A quick way to get a count would be search the log files for keywords such as bot, bots, robot, crawler, spider looking for all legitimate bots who link to web pages explaining how they crawl. Or search for robots.txt i'm sure most bots would look at it even malicious ones to see where you don't want them to crawl. –  Anagio Aug 7 '12 at 10:54

This week our company (Incapsula) launched Botopedia.org - a Community-Sourced bot directory. It's 100% free and open for all and you can use it to find a complete user-agent list for all bots you`ll want to look up.

As for indentification methods, I want to refer you to this discussion in Security.Stackexchange which covers different methods of bot identification (i.e. JS challenge, Method check, robot.txt access and more).

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.