|
#1
|
|||
|
|||
|
Google Base Opens To Submissions
It's offical. Google Base , Google's new system for accepting submissions of anything, is now open. Google Base Live, Accepting Content in SearchDay today has a full rundown, Gary shares thoughts in Poking At Google Base and if you need more, Memeorandum has an excellent recap of stories. What do you think?
|
|
#2
|
|||
|
|||
|
no luck for me ... I encountered Google Base: Error
"Oops! There was an error with the page you requested. Please check back later." |
|
#3
|
||||
|
||||
|
..all your base belong to us...
<mod hat off>
Google analytics, Orkut, Gmail, Adsense, Adwords, Google Earth, Google desktop, Google toolbar - what an amazing portfolio of methods to collect user data, and then serve advertising to those users. With all that data - Google will have the most targeted advertising systems ever offered. Now - 'All your base' really do belong to Google..... <tin foil hat on; mod hat on / > ![]() |
|
#4
|
|||
|
|||
|
Just listed my 'Web services' onto Google Base. It's good to see the other 'smaller' companies (SEO (with an S), web design and development) doing likewise
It took approx. 45mins to clear before it got listed. I have already emailed my clients before it breaks in tonights news here in the UK! Nice location map (Google Maps) and upload image feature (10 images max)...still playing with GB just trying to think what else I can get on there!! Last edited by Andy1969 : 11-16-2005 at 08:43 AM. |
|
#5
|
||||
|
||||
|
Great offer/demand base interface.
As many query mechanisms, this one can be customized. To return top 100 results for car sorted by relevance, try http://base.google.com/base/search?n...coring=r&q=car Change scoring=l to sort by oldest post date scoring=ld to sort by most recent post date Great venue for online classifieds and for mining seo's "Classified Optimization Services" Orion |
|
#6
|
|||
|
|||
|
From Danny's article:
Quote:
![]() |
|
#7
|
|||
|
|||
|
google base is not working
Quote:
http://base.google.com Oops! There was an error with the page you requested. Please check back later. |
|
#8
|
|||
|
|||
|
Are Google risking Paid Search revenue???
Quote:
|
|
#9
|
||||
|
||||
|
Quote:
Can we expect this on Spanish version soon? |
|
#10
|
|||
|
|||
|
It's interesting. Am I mistaken, or do people need to have a Google account to actually search the database (i.e. not open to the general public)? If it needs an account, then I think it's usefulness is very limited, but I'd think it would find its way onto Google's homepage as a tab, and then it would be very useful to list things in.
Last edited by PhilC : 11-19-2005 at 10:04 PM. |
|
#11
|
|||
|
|||
|
Quote:
You do not need an account to search the Google Base, if you find yourself logged in, just log out Andy |
|
#12
|
||||
|
||||
|
I'm loving Google Base! It is a more appropriate place to upload our course schedules than Froogle and it's nice to be able to search the various doc categories. Being able to add an image is a cool feature.
__________________
[COLOR=RoyalBlue][SIZE=2]Kalena Jordan[/SIZE][/COLOR] [FONT=Verdana]Search Engine College[/FONT] |
|
#14
|
|||
|
|||
|
But it's normal for good ideas to be emulated.
|
|
#15
|
|||
|
|||
|
There seems to be a problem with Google Base and bulk uploads. A number of people are saying that Google finds their URLs to be invalid, but the URLs look fine.
I uploaded a file with 4248 URLs in it, and they were all declared invalid. As a test I submitted one of the URLs by hand, and it was published ok. So there seems to be a problem at Google's end. I'm going to try another test and remove all but 10 of the URLs, but I have to wait, as they only allow one bulk upload per hour. added That didn't work. All 10 URLs were declared invalid. Now I'm going to try it with just one URL. added Having just one item in the file didn't work either. The URL that Google says is invalid is:- http://www.forthegoodtimes.co.uk/cam.../abergele.html I wonder what valid URLs look like ![]() Last edited by PhilC : 11-22-2005 at 12:49 PM. Reason: additions |
|
#16
|
|||
|
|||
|
There's only me left in here now
![]() There's something that I don't like with the Base. Once you've registered a filename, you can't get rid of it. You can delete the file itself, but you can't delete the filename's registration. Neither can you re-use it after deleting its associated file. For my tests, I had to keep using new filenames, and when the problem is solved, I want to use the orginal one, but I can't have it. I've got it registered, but I can't use it. I think that must be an oversight on Google's part. I think that, when a file is deleted, its filename registration should also be deleted, or at least it should be on offer to use again. |
|
#17
|
|||
|
|||
|
You can tell you're geting old when you start talking to yourself - as I'm doing in this thread
I've tried around 8 experiments to get GoogleBase to accept any bulk upload items at all, and they've all failed with "invalid URL" errors throughout. Their system has problems. I'll explain what the error means. You can upload a file that contains up to 100,000 items for publishing. The file must be in one of several formats, and I chose the tab-delimited text format (.txt). The first line of the file must contain the field names of the data you are providing for each item. It must include "title", for instance. I used title, description, label and link (label is effectively keywords, and link is the full URL of the webpage, including "http://"). They are all Google defined field names, and they were tab-delimited, so there was nothing wrong there. Each item in the file is a one-line, tab-delimted list of the data for each of those field names. The "invalid URL"s are the ones in the link field. Google even shows you each of the URLs, and tells you what line in the file it is on. Every URL in every file that I tried was declared as invalid. And yet when I submitted them by hand, as single items, they always went through ok without any errors. I checked the files down to the hex level to make sure that the tabs were where they were supposed to be, and that no spurious invisibale characters had found their way in. There was nothing wrong with them, so there must be problems at Google's end. If you try to upload a bulk file, and you get problems, you know that it may not be your fault. There are a number of people in Google groups who are having the same "invalid URL" problems when the URLs are perfectly valid. On the other hand, there are people who are getting bulk files through ok. You only need to search the Base for 'hotels', and you'll see stacks of them. Personally, I've given up on it until Google fixes it. |
|
#18
|
|||
|
|||
|
Hi Phil
So you don't feel too alone I have been reading your posts with interest. I haven't had to do a bulk upload yet as my clients are still trying to see if Google Base is for them. But as soon as I do I'll be back here seeing if you have (sorry if Google has) fixed the problem. Keep it up...good post cheers. |
|
#19
|
|||
|
|||
|
cheers Andy
![]() I read some more at Google Groups and there are various faults with bulk uploads, and not just the invalid URLs problem. Some people even tried uploading Google's own example files, and they didn't work properly either. They got published ok, but some of the data just disappeared. I don't know why any clients, or anyone, would hold back from using it. Maybe I haven't thought it through, but it just looks like a nice freebie to me - instant publically available rankings without doing anything to the site - it's not a webpage listing system, and items that are webpages aren't spidered. The system scans the "description" field to sort out the rankings, and that field can be up to 64k in size. It must use other fields as well, such as the title and other attributes, but the description field is particularly open for optimizing. To be honest, I don't understand why they are doing it as they are. It's too open to spam - except it isn't spam if it's allowed. They've got a lot of work to do on it if they want it to end up as I imagine they envisaged, rather than flooded with hundreds and thousands of pages from individual sites - as it is now. At the moment, they encourage that sort of flooding by allowing bulk uploads of up to 100,000 items (webpages for us) per bulk upload file, and multiple files are allowed. It is a publically searchable utility that bases its rankings solely on the content of the data provided. Items are indexed, and rankings seen, pretty much straight away, and the data can be altered at any time to improve the rankings. That's how some major (now failed) search engines used to be. |
|
#20
|
|||
|
|||
|
Google did a quick fix and I got the bulk upload through and the webpages published ok. Their fix was to change the short "Invalid URL" message into a lengthier one that states some of the reasons why the URL may be invalid. One of the reasons is that the URL does not belong to the site in my account, and that was the reason why all my bulk upload URLs were invalid. I changed the site in my account and they were all published ok.
I've deleted them all now - I didn't like having the top 780 results for a popular searchterm - especially as they still only show the top 1000. Some system, huh? But it's beta. But it raises an issue, imo. Anyone can publish any webpage from any other site and give it a derogatory title, description, labels (keywords/phrases), and other attributes, and get it ranked for any query that takes their fancy. I could publish the homepage of searchenginewatch, for instance, and describe it as "The worst seo resource on the web - run by a person who is <insults and lies>". I could also publish an individual's website, or contact page, and have it show in the results for anything I like. I'm not even going to write the possibilities that have sprung to mind. Because the system doesn't crawl webpages, the rankings are based on the data provided, so it is possible to tweak the data until the unscrupulous pages are ranked at the top. And there is only a few minutes waiting time between tweaking the data and seeing the effect in the rankings. I don't like that 'open' aspect of the system at all. Imo, publishing webpages should not be possible without the same sort of website verification that Google requires for their additional Sitemaps data. Last edited by PhilC : 11-26-2005 at 01:37 PM. Reason: addition |
![]() |
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | |
|
|