Personalisation

By: Boris Rousseau

Have you ever found that searching the world wide web can be quite frustrating? Anyone who surfs the web often has more than likely been confronted with the problem of being flooded with masses of information of low relevance and having to pick through it for the relevant pieces. Now that users can access the Internet from from various devices not only the pc, such as laptops, Personal Digital Assistants (PDAs) or WAP-enabled mobile phones the problem is compounded further because not all information on the Internet is viewable on all devices.

Consider for example, John - a user who already has a degree in french and he is looking for a post-graduate course in french. If he enters a search request for "courses in french" using any of the current search engines, John's previous qualification will not be taken into account and therefore he will get a number of results, which are not relevant to him. Furthermore, some of this information cannot be displayed properly on his pc.

So, how can John get the information he is interested in, tailored to his device capabilities and his needs? The answer to this problem is Personalisation.

The idea behind personalisation is to get very accurate searches results by maintaining a record of the user's interests, preferences, device capabilities, etc. This user record (or user profile) is used to filter information tailored to the user's specific needs. Typically a user profile would contain some dynamic information such as device capabilities discovered "on the fly" by a dynamic detection process, and action history recording the user's previous searches and content selection. A combination of this dynamic with static (contact details, interest, preferences) information will narrow search results and make the whole search process a lot simpler.

Personalisation is not only applicable to the Internet but also to any Telecom service that requires customisation and adaptation to user requirements. Such service could be a mobile service where information has to be adapted to the particular characteristics of the device.

Let's consider another scenario. Lisa, a student in astrophysics, has just watched a report about black holes on her interactive TV and wants to find out more. If she enters the keyword "black hole" into a search engine she will receive lots of information on the subject but it is also likely that she will only be interested in advanced information more relevant to her background. The system implemented in GUARDIANS a research project at the TSSG (www.ist-guardians.tv/) will take this information into account and provide the possibility to save relevant results. For instance, if Lisa is currently accessing the system from her PC and finds out that there is a relevant report, she can save this information to her pc for later. The fact that Lisa is accessing the system from various devices such as her home PC, her laptop, her Personal Digital Assistant (PDA) or a WAP-enabled mobile phone is also significant. For example, while travelling, she might decide to search for a video footage of the formation of a black hole from her PDA and using the Guardians system she can download it directly to her home PC to view on her return.

As the number of users connected to the Internet and the number of available resources continues to grow, the personalisation of information will become even more crucial. It is expected that user requirements will be the cornerstone driving future Internet and Telecom services.