Stop indexing of PART of a Web page?

[Silver Surfer]

Is it possible to stop Google (and others) from indexing part of a Web page? In particular, due to privacy concerns, i want to hide a few names in a phone directory from google without de-indexing the whole page with robots.txt-- can this be done somehow?

Thanks!

[PhilC]

Nope, it can't be done. Danny just posted in another thread that Yahoo! is looking at introducing a tag that will do just that though, so it may not be too far away.

An alternative would be to display the stuff you don't want indexed in an inline frame. Google doesn't follow inline frame sources although other engines might. To be on the safe side, you could use the noindex meta tag or the robots.txt file to disallow the indexing of the inline frame page.

[Chris_D]

i want to hide a few names in a phone directory from google

Put the names in an image - i.e. a 'picture of words' - rather than actual words.

Search engines can't read pictures of words....

[PhilC]

Good idea, Chris. Why didn't I think of that

[ohcho]

Also use <IFRAME> and/or <WBR> so that it won't be able to searched. Of course IFRAME text with the followwing meta tag;

<meta name="robots" content="noindex,nofollow">

[softplus]

Why not do it on the server side? I'm assuming it will only effect a few pages, you could easily turn them into php or asp pages (if they aren't already) and return different names (say "joe schmoe" instead of "micheal mcnumen"). Of course this goes under cloaking, but so do any other methods you use to give search engines different content than human visitors.

But anyway, if you are giving visitors "free access" to that page with the full content, where is that any different than having the search engine crawl/index/archive the real content?

Perhaps it would be a good idea to add a "preview" page to these confidential pages, combined with some sort of "free login" or captcha to access the real pages (and mark the real pages with robots=none + put them into the robots.txt). That way your preview will get indexed but the real contents won't. (oops, doorway page ;-))

Another possibility would be to use javascript to display the names... or use CSS to hide a fake name and display the real one (split up into different spans). I don't know about using an image -- Google indexes those already (though you can turn that off for your site in the robots.txt file).

Lots of possibilties - can you let us know what you used in the end and why? Thanks!

Cheers

[Silver Surfer]

Thanks so much for suggestions. I went with the <IFRAME>, which seems to work fine on my server and on all three major browsers. I had to tweak the params to get it to look right. Here is what I ended up with (to insert some plain text):
<iframe src=inc.htm frameborder=0 align=bottom
scrolling=no height=15px width=100px marginwidth=0px marginheight=0></iframe>

Picture was not right for this application, because it is a directory with a whole list, would have been too much work/impossible to get it to look right on all browsers.

Thanks!

[ohcho]

As a matter of fact, I use it. So I suggested. The reason I started hiding was that spammers, I mean real spammers, search using certain information to verify email status and match identities.