Using Bing Web Search API, I need to filter results only for my domain, example query:
https://api.cognitive.microsoft.com/bing/v7.0/search?q=site:mysite.com+myquery
But in results I received not only mysite.com results but also from sites like wikipedia and others.
How I can search result only for my domain?
Bing Custom Search not work for me because I have more than 10k transactions
Your website is probably not known or indexed by Bing. Since you use Bing Search API, not a custom indexing service or search across a sitemap.
The actual Bing website will need to be able to find your site.
Since it doesn't, this triggers the default behavior of returning as relevant as possible results.
This behavior is valid for urls such as the following:
https://api.cognitive.microsoft.com/bing/v7.0/search?q=microsoft+site:notAnIndexedWebsite.com
Formatting wise, there is multiple options, as seen here.
None of it is a problem in this case.
You can try Bing custom search if it helps (at https://www.customsearch.ai/), especially as it is now in GA. It also provides option to crawl your pages in Bing index through webmaster, if they are not crawled already. This should make sure that you get results only from your website.
Related
I'm currently using woocommerce and by chance when looking through google search console I noticed well over 15,000 pages had been indexed. Instantly raised a concern because I know I should have no more than 400 actual pages.
After looking into it, I noticed that absolutely every possible parameters "variations, grid styles, shipping methods" etc is being indexed causing what is 400 into 15,000 variations.
Does this have an affect on google ranking showing 15,000 pages when really there are only 400.
I can not find a single resource that explains whether google indexing so many variations has a positive or negative impact on google rankings.
Finally how to prevent google or any other search engines from indexing url's with parameters. I've seen advice using robot.txt but no recommendation of what standard woocommerce filters should be excluded or if that's a bad idea?
I have a feeling I am losing alot of link juice by having so many indexed pages with parameters?
WordPress by default expects that you wish to share everything because why not? Surely if you have 15,000 items that are index-able surely you'd be more mad if your site decided on a whim to hide parts of that without you knowing. It is understandable that this would be confusing though, and quite rightly it is always better to have control over which parts of your site gets seen by search engines.
Anyway there are plugins which allow you to customise which parts of a website are visible or hidden to search engines by providing a mechanism for generating an XML site map (an index of which pages should be crawl-able). A very common plugin I see people use is called Yoast SEO. Another one I have had some success with is called Google XML Sitemaps, you may find another to work better for you, but that should give an idea of what you need.
Also if you are not getting visited often by google bot then take the sitemap from one of those types of plugins and submit via the google console to help google better understand your site.
I need some solution to make url as on image shown. I think regular expressions don't work in that field, however it announced that regex works in google analytics.
Regards, Dmitry.
Regex in Google Analytics is intended to be used in features that affect a broad range of pages or fields, like filters, advanced segments, etc.
The web page field from your screenshot is looking for a static URL. The content experiments are designed to run variations on a single page.
Been using internal site search with Google Analytics and while I love the ability to see what my users are searching for, I am having a really hard time figuring out what search terms lead to which pages.
When I search on both the nextPagePath and searchKeyword dimensions while filtering on the search results page at the current path, the nextPagePath is always the search results page even when I know it shouldn't be (when tracking my own obscure searches). The same goes for using the searchDestinationPage dimension. I can't get any data that shows a jump from a search results page to another page on the site.
Here's a cleaned up example of my api query.
dimensions=ga:searchKeyword,ga:nextPagePath&metrics=ga:pageviews&filters=ga:previousPagePath=#dosearch
When I use the standard Analytics UI and look at the Destination Pages list under Content->Site Search->Destination pages, I only have 25 or so, all of which are just the variations on the base search-result page URL.
Do I need additional tracking code on my search results pages? Custom variables? A different query through the API?
I can see the tracking requests going out from both the search results and the pages selected from the results.
I found a couple of questions in the Analytics forums that ask this same question, but none of them had anything resembling a working solution.
I would bet you are not using the proper dimensions in the API
See https://developers.google.com/analytics/devguides/reporting/core/dimsmets/internalsearch
ga:searchDestinationPage is probably what you wanted when using ga:nextPagePath
I've an ASP.NET page with a textbox and a option from user of the following choices: Wikipedia, Google, Dictionary.com, Flickr, Google images.
The user enters a word(s) in the textbox and selects a choice among the following.
Depending on the choice select by the user I wish to return the following.
Wikipedia: Return the content and link to the page corresponding to the topic about the word.
Google: Return the top 10 results of google search for this word.
Flickr: Return a few images atmost 10 images from flickr search
GoogleImage: Return a few images from google image search.
Dictionary: Return the meaning of the word.
How can I do that?
Since you are wanting to do some processing on the results prior to displaying them, your best bet is probably to invoke a web request on the server to fetch your results as RSS or some other parsable XML format.
So first up, we have Wikipedia, which has API support for open search, and queries with XML or JSON output. You can get the details of the API by going to: http://en.wikipedia.org/w/api.php
I would think either the query action, or opensearch action would be what you want.
Right, now there is Google, which supports search results as RSS through their Active Search feature. The link takes you to the main page where you can build the query, at which point it should be easy to drop in your search terms. There is also the Google Search AJAX API, which you can find out about here (See the "Flash and other Non-Javascript Environments" section for building the URLs directly. I believe this option should give you access to Google Image results as well.
For Flickr, have a look at this App Garden page. There are several output formats available to choose from.
I wasn't able to find anything real solid on getting results from Dictionary.com, but it does appear that they have an API. You might be able to dig through google and find some references on how to get search results as XML or JSON. There are also several other Dictionary sites which may have more information about their APIs. While searching I managed to find this SO question about word lookup from google dictionary.
Hope this helps.
Have an iframe within your page, and then set the src of the frame to the appropriate query string that you craft from the user's input.
This can be done from javascript within the page, in response to the user selecting something in the 'choice' dropdown. You can have the appropriate urls already embedded in the javascript (as variables), and just substitute in the user's input.
I'm doing custom-rolled view tracking on my website, and I just realize that I totally forgot about search bots hitting the pages. How do I filter out that traffic from my view tracking?
Look at the user-agents. It might seem logical to blacklist, that is filter out all the strings that contain "Googlebot" or other known search engine bots, but there are so many of them, it could well be easiest to just to whitelist: log visitors using a known browser.
Another approach would be to use some JavaScript to do the actual logging (like Google Analytics does). Bots won't load the JS and so won't count toward your statistics. You can also do a lot more detailed logging this way because you can see exactly (down to the pixel - if you want) which links were clicked.
You can check the user agent: here there is a nice list.
Or you could cross-check with the hits on robots.txt, since all the spiders should read that first and users usually don't.