| 
  • If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • Finally, you can manage your Google Docs, uploads, and email attachments (plus Dropbox and Slack files) in one convenient place. Claim a free account, and in less than 2 minutes, Dokkio (from the makers of PBworks) can automatically organize your content for you.

View
 

Search Engine reflective notes

Page history last edited by PBworks 12 years, 3 months ago
Search Engine reflective notes

 


Posted by: Tony 21.11.08

 

Following the trial of the word processing module that ended last week, a number of important messages were received about the effectiveness of the customised search engines to identify relevant learning content. There was some uncertainty about the ability of the customised search engines to perform better than general search engines. An initial comparison by one learner appeared to indicate parity of performance. This was not the experience of other participants, but an outcome of the consultation will be a structured comparative evaluation with a view to optimising cse performance.

Another level of uncertainty related to the way in which the search results should be interpreted and used. Some work needs to be carried out on how search results are used in informal search scenarios (which will be the majority of circumstances) and to integrate effective practice into learning practice.

 

Posted by: Tony 13.10.08

 

The second ICT module covering spreadsheets is now in draft form and may be accessed here. It follows the same format but has a different search engine associated with each node. If this approach works, it will prove a very efficient way of developing new modules. The basic competence map structure is now being used as a template for document design and production and the embedded search engines are modified appropriately each time. The new draft module took less than four hours to produce including the new search engine, competence map and Wiki page.

 

Posted by: Tony 19.09.08

 

The team has been discussing ways in which control is given to the learners, not only to follow their own discovery learning pathways through the search engine results, but also to refine the focus of the search engine and to customise the search to their own particular interests and, as this is ICT, to their own particular mix of hardware and software. The current CSE for ICT1 has been set up as a generic search engine for Word Processing resources. Each of the concept map nodes enters a different search string relevant to the competence involved. This brings up a set of resources that represent the learners' starting point for that particular topic.

The use of 'Google Marker' allows the learners to refine the search engine performance by identifying web sites that are particularly effective at meeting their skills needs. This adds the new URL to the CSE Annotations list and hence steers future searches in that direction. There will need to be an overarching management/moderation process for this, but this is the first step towards a peer-rated set of topic focussed resources that keeps the resources current and relevant.

When the search results are presented, the Google page shows the original generic search string for that topic/competence at the top. The learners will be guided in the choice and use of additional search terms to home in on resources relevant to their particular word processing application.

 

Posted by: Tony 24.08.08

 

There are a number of other possible options available for conducting searches based on networks of bookmarks. Delicious itself provides a search function that will selectively search your bookmarks, your network's bookmarks or everyone's bookmarks for results. Nsyght will also use your bookmarks and your network's bookmarks as the basis for its search. Other, more specific content sharing services are also available such as Twine and Lijit.

From the volume of development activity out there at the moment this is clearly seen as a core web 2.0 area of functionality. As the services are refined and improved, the benefits for this kind of discovery learning approach will follow. Given the limited scope and timescale of the eTutor project, it will not be possible to fully explore all the various options available. However, a current overview and scoping will be provided as part of the project dissemination and reporting.

 

Posted by: Tony 22.08.08

 

In addition to the development described below, a version of the Google CSE hosted by Google, but locally optimisable (if there is such a word) has been constructed and is ready for testing within the pedagogic and learner support models. The Ask.com version has also been completed for comparative testing purposes.

 

Posted by: Tony 20.08.08

 

Progress was made towards this target today by converting the Google hosted CSE to a locally hosted CSE based on this website. See the linked CSE on the Experimental page of this Wiki. Both the CSE Definition and the associated Annotations are files on this site that are accessible for updating locally. The Annotations file contains all the URLs used to focus on specific search objectives. The next step in the development process will be to devise a method of automatically updating the file based on learner ranked learning resource sites.

 

Posted by: Tony 19.08.08

 

I think I've now reached the limit of ASK.com utility in this application. Although the effectiveness in identifying relevant sites is excellent through targetted tagging, and the opportunities to refine the search on the fly support discovery learning, the manual setting of tags is when constructing the module concept map is laborious and would require some form of automated updating to ensure optimal performance. I will be looking at the possibilities for automation later in the project.

 

Automation is also the issue with Google customised search engines (CSEs). I have spent some time manually simulating automatic CSE optimisation by ranking and tagging URLs identified by the CSE and then updating the URL and tag listings using the CSE editor. I believe it has a great deal of potential following my initial experiments. I have created a curriculum model for Word Processing that I have applied the approach to and the results are very encouraging.

 

What needs to be done now is to complete the CSE optimistion model shown below by finding a way to automatically update the CSE with ranked URLs and tags.

 

 

This will be the next task for the project team.

 

Posted by: Tony 10.07.08

 

Experimentation to date has shown that the Ask.com functionality lends itself better to the discovery learning approach than customised Googles. It is also much easier to implement, refine and manage. The entry to the process is through a hyperlink from each of the learning outcomes on the module mind map. The link can be pre-loaded with search tags and I am experimenting with the most effective combination for each learning outcome. Concept Maps are now also being evaluated for use in this application.

 

Posted by: Tony 15.05.08

 

The learning activity chosen was Word Exercise 4 from the ICT module delivered as part of the WeTN Cert HE in e-Commerce. This activity resulted in the production of 2 forms using formatted tables in a Word document. A Google Customised Search Engine was created to search for resources covering the creation and formatting of tables. This may be viewed on the Customised search engines page.

 

Posted by: Tony 14.05.08

 

What is required is a search engine that will provide links to resources that can be directly used to develop the skills and knowledge that fully cover the learning outcome selected on the mind map. There are two different approaches being evaluated in the current phase of development. The first is to experiment with the use of Google Custom Search to 'tune' the search response to achieve the required specificity. The second is to use the Ask.com search engine which includes user-based 'search narrowing' capacity to tune the response.

 

The approach for the experimentation will be to select a specific learning outcome from the ICT_1 mind map and to create a learning activity that will require the totality of the skills and knowledge to be demonstrated for the activity to be successfully completed. The process of completing the activity will then be carried out using only the information from sources returned by the web searches.

 

Comments (0)

You don't have permission to comment on this page.