Week 10

For my final blog of my COM425 E-Commerce module I will be discussing web personalisation. and my personal views on it.

Personally, I am for web personalisation..  I am for this because I believe that it gives the user a chance to view a page that is specific to them and includes relevant information to the user.

http://www.archimuse.com/mw2004/papers/bowen/bowen-fig1-inline.gif

Web personalisation is done in 3 stages:

• Input
• Algorithm
• Output

The input is the user viewing or purchasing products or information, the algorithm is done by collaborative filtering and simple data mining to show which products and information should be displayed.  The output is the relevant products and information being shown.

Personalisation works by using collaborative filtering, finding out information about the user and displaying information and products that the webpage believes to be relevant and useful to the user.  A simple example of this is on e-commerce sites Amazon (www.amazon.co.uk) and Play (www.play.com), where products are displayed and recommended to the user.  These results are dependant on what the user has searched or purchased in the past.  I admit that this can also be a downside as the products etc that are displayed are not always relevant to the user.

www.amazon.co.uk

The image above shows products that have been recommended for me by Amazon.  This is shows both good and bad web personalisation as I would be interested in the 1st and 3rd games shown.  It shows bad personalisation as I would not be recommended in purchasing the 2nd game shown.

Another example of bad web personalisation would be when I go to Amazon to buy a product different from my previous purchases, and I am constantly being recommended with irrelevant products.

With the web moving into its next generation, Web 3.0, personalisation is going to be a ever present form of displaying information and data to the user.  Many experts believe that the web will become like a personal assistant to the user.  The browser will do this by learning from the users searches and purchases, allowing the user to be less specific with their searches.

Advertisement