Personalization for the Mobile Web: A Position Paper

Personalization for the Mobile Web: A Position Paper

Michael J. Pazzani
Department of Information and Computer Science,
University of California, Irvine
Irvine, CA 92697-3425

The wireless web offers the potential of access to vital information anywhere at any time. However, hand-held devices with small screens connected to the web with slow and expensive network connections offer new challenges. In particular, it is difficult on such devices for a user to scan through lists of e-mail messages, restaurant listings or news stories to find personally relevant information. Services that can personalize information for the user based on the user's location, current task, and a profile of the user's interest can make the wireless web live up to its potential.

Personalization via User Customization

The dominant model for personalization on the wireless web is illustrated by services such as MyYahoo that allow a user to select check boxes indicating interests. In these services, an account is created on the Internet, and the user selects check boxes to determine what type of content to send. There are several problems with this type of personalization. First, due to the number of options available, the user cannot use a wireless device such as a phone to display and edit the options. The complexity of setting up an account on the web and associating that account with a wireless device may limit the audience for personalization of this type. In some countries, there are more cell phone users than Internet users limiting the market for such services. Second, the options tend to be too coarse-grained. For example, one can ask for sports scores in MyYahoo, but there is currently no way to indicate a particular sport, or a particular team to follow. One could image creating such categories for Sports, but in other areas, such as world news, it would be difficult to anticipate all possible distinctions, since some users might want to follow news on East Timor, and others Chechnya. Third, such profiles can be used to filter news but not to prioritize it. It is unlikely that the same number of stories that fit on a user's screen will meet the profile each day and some way must be found to determine which stories to send first.

Keyword-based systems are an alternative used by some services that are moving into wireless web from the content delivery to pagers. They rely on the user creating a profile that consists of a number of keywords indicating the type of content they would like to receive. Due to the complexity of text input on the cell phones, these services also require a user to create an account or profile on the Internet that determines what content will be delivered. We believe few users will be able to construct profiles with such keywords. Like dialog boxes, keyword-based systems do not help in determining the order in which content is sent to users.

Adaptive Personalization

The above approaches are pursuing an approach where constructing a user profile places the burden of personalization on the user. An alternative approach is to use artificial intelligence or statistical techniques to automatically construct a profile of the user interests. We advocate such a strategy. It is important that adaptive systems learn from a small number of examples and adapt quickly to changing user interests. To be truly useful in a mobile context, user interest must be inferred implicitly from the actions rather than requiring explicit ratings of content by the user. For example, the more a user reads of an e-mail message or news story, the more a user is interested in that story. Explicit feedback of user interest would require a user interface consuming screen real estate and additional transmission of data.

There are two basic approaches that may be used to infer profiles for making recommendations to users: collaborative filtering and content-based filtering. Collaborative filtering monitors the behavior of all users and tries to find users with similar tastes. It then recommends items to an individual if similar individuals like it. Systems by companies such as FireFly and NetPerceptions have been used to recommend CDs, movies, and books. This approach requires many users to rate an item and users to rate many items before it can reliably make recommendations. Collaborative filtering is appropriate for personalizing location-based services such as restaurant recommendation on the web. Applications such as recommending nearby restaurants and retail outlets can be enhanced adaptive personalization technology so that the best nearby options are presented rather than simply the closest options. However, it not appropriate to e-mail and unlikely to work well for changing events such as news delivery.

Content-based recommendation creates a statistical profile of the user's interests in terms of the words and phrases that distinguish items of interest to the user from other items. It is appropriate for news, e-mail and restaurant recommendation other filtering where a text or structured description of content is available. We have developed such a system at the University of California, Irvine and evaluated the system on a group of over 3000 users on news delivery. Our results show an increase in news readership by over 40% when headlines are sent in a personalized order to user.

The adaptive personalization approach with implicit feedback has two important advantages. First, from a user's viewpoint, it is easier to use. The user just uses the system and a profile is automatically constructed. The user need not be told that there is an adaptive server since it requires no user configuration. That is, it is a marketing decision to decide whether to advertise this and not a technical necessity. Second, it allows for additional content-based criteria to be used in determining which stories to send to a user. In particular, in addition to sending relevant news stories or e-mail messages, we also desire to send a diverse group of items. This can be achieved by measuring the similarities among current items and making sure that the items on a screen are relevant and dissimilar to one another. By ensuring some diversity among the items, the most of the bandwidth is devoted to items of known interest to the user, but some bandwidth is reserved for items that are novel and can then determine whether the user is interested in following a new area.

Personalized Information Access for the Visually Impaired

Many visually impaired Web users access content through the user of screen readers that convert text on the screen to audible speech. The constraint imposed by the screen reader are not that are not dissimilar to those imposed by the Mobile Web user. In particular, information is delivered slowly to the user. In this situation, it is equally important that relevant content be presented first and that the personalization be easy to use. We have created a client for the Jaws screen reader using the adaptive personalization approach advocated here.


We have argued that adaptive personalization is an important component of the mobile web. To fully realize the goal of having access to important information anywhere, it is necessary to understand what is important to the user and use that to order the transmission and presentation of information.


Prototypes of adaptive personalization approach discussed here are available for use in a variety of formats:
  • Web Clipping for the Palm VII: Download PQA
  • HDML
  • WML
  • HTML formated for PDQ cell phones, RIM pagers, Palm and Windows CE browsers
  • HTML formated for the Jaws Screen Reader

Michael Pazzani
Department of Information and Computer Science
University of California, Irvine
Irvine, CA 92697-3425