Archive for the ‘Search’ Category

Step 1

You need to send in a request (email) to the ICT service desk.  Please

1. Give the url of the website that you would like to have the search filtered on (i.e. www.imperial.ac.uk/webguide) and
2. Make the email for the attention of the BS web support group
3. Also if you want it placed in a certain place on a page then please indicate the url of the page and the position within that page and we will insert this for you

With this request, a ‘collection’ will be created in the Google search application which filters the results for this nominated website.

Step 2

Note: This step is ONLY required if you have not indicated where you want the filtered search appearing (point 3 above).

A html portlet needs to be added to the page and the following code inserted:

<form action=”http://www.imperial.ac.uk/search/results” method=”post”>
<label id=”hidelabel” for=”searchwords”>Search for:</label>&nbsp;  
<input maxlength=”1000″ name=”searchbox” type=”text” value=”Search ICT” />&nbsp;   
<input id=”quicknav” title=”Search” alt=”Search” src=”/pls/portallive/docs/1/9749699.JPG” type=”image” value=”search” />&nbsp;  
<input name=”collection” type=”hidden” value=”XX” />
<input name=”sr” type=”hidden” value=”0″ />
<input name=”secure” type=”hidden” value=”1″ />
</form>

Where XX is the name of the website (i.e. webguide).

What is Meta Data?

Search engines such as Google use page titles and the text contained in web pages when building their indexes and ranking search results. Some search engines will also use ‘meta data’ if it is present on a page.

The CMS templates automatically generate some meta data for all pages, so you do not need to manually add meta data at all. However, if you want to try to improve the ranking of some particularly important pages in your site, it could be worth your while spending some time and effort on meta data.

Adding keyword meta data to a page may help to improve the ranking of a page by some search engines by attaching additional words which do not appear in the actual text of the page e.g. a page about ‘training’ could have meta data including the keyword ’teaching’. The page might then be returned in the search results for a user searching for ‘teaching’ even though that word does not actually appear in the text for that page.

Another type of meta data is the description which is displayed in the list of search results by some search engines. An interesting description might encourage an internet user to visit your page rather than others returned in the results list.

Keywords and Description are added in the Edit content view of a content item, and apply to the whole page. The boxes to add them appear below the main editing window.

Notes on meta data

• The keywords and description form part of the meta data of the page and aid search engines (e.g. Google) in ranking pages.
• Keywords format: separate the words by a comma and a space (i.e. earth sciences, complex system modelling, cosmic).
• Description for new pages: if you do not fill in the description, the header1 on this page (if there is one) will be automatically inserted as the description.
• Sub-pages inherit keywords from their parent page when they are created but these can be changed/removed/added to as required.

Google Search Appliance was chosen by the College in the Summer of 2009 as the best suited search engine for Imperial’s requirements. ICT operates the College search engine.

What is indexed?

The Google Search Appliance is indexing appropriate College web content (HTML, word documents, pdfs, blogs etc), including the Content Management System (CMS), Medicine CMS, sites on the Windows and Linux web farms, the Imperial College Union site, etc.

The index includes publicly available content as well as some permission based content, specifically permission based pages in the CMS that anyone in the College can log in to see. When a user search results in permission based page results, users will be asked for their username and password before they can view the content. Any content restricted to a smaller subset of College members will not appear in the search index.

No. The license for Google use limits the number of pages we can index and we are prioritising our primary sites.

Yes – contact the Service Desk for further information

Where a page is password protected on the College website, Google will display brief summary information, but will demand a College log in when the result is clicked on.  Google will only index password protected content which is accessible to all College members.  Any content restricted to a smaller subset of College members will not appear in the search index.

It is possible to exclude your page from the College search results if you have a strong reason for doing so. Remember, many users only use Search to find content, so excluding your page from the listing may mean no-one can find it.  If you do want to discuss excluding your page, please contact the Service Desk.

• Follow Imperial’s web standards.  Search engines find it easier to access properly constructed, validated and accessible code.
• Think about your URLs – make sure you include a clear keyword in the page name
• Make sure each page has a clear Heading 1 <h1> which includes a content keyword
• Ensure your page is linked to from other pages (e.g. from other pages within your department site, and perhaps within the Admin and Services A-Z or the A-Z of Departments and Centres)

If your pages are still not being ranked highly for relevant searches, please contact the Service Desk

Search results are sorted by relevancy according to Google algorithms, using text-matching techniques to find pages that are both important and relevant to your search. For example, Imperial’s Google search ranks pages according to how near each other your search terms appear.

The College search engine is regularly refined and improved.  Please report any inaccurate search results to the Service Desk.