Privacy vs. Personalization

Libraries have had little success with creating personalized, individual services.

Recent environmental changes make personalization achievable. The concepts of user as contributor and rich user experiences are key aspects of Web 2.0 and both are tied to personalization of the web adventure. User as contributor enables users to share their content with others, whether it is research, photos, writing, or videos.

Additionally, the concept of a rich user experience demands services and encounters that are geared to an individual user. If we want libraries to participate in Web 2.0, we have to rethink our perspectives and policies regarding privacy issues.

Nearly all user experiences on the web are personalized. Amazon and Netflix market new content to users based on what they have previously consumed. At Amazon, this takes the form of “Customers who bought this item also bought” suggestions and the “recommended for you” section. At Netflix, it is the “Movies You'll Love...” tab when you login. This personalized industry has fed the growth of the long tail, since making the entire realm of content visible drives hits and outlying items.

My Yahoo was one of the earliest examples of personalization on the web, but Web 2.0 personalization goes much further. In fact, tools that allow users to tag and collect their preferred materials are also personalization tools of a sort, because they help “personalize” the way in which users interact with the web by keeping content they have selected close at hand. The most notable examples are del.ico.us, LibraryThing, and CiteULike (see Product Pipeline, netConnect, Summer and Fall 2006 and Winter 2007, LJ 7/06, LJ 10/15/06, and LJ 1/07). While Netvibes and PageFlakes are early examples of what users can put together, Yahoo Pipes takes mashups and personalization to a whole new level.

Web 2.0 is participatory in nature, and most users who are employing its tools are making a choice to reveal particular bits of information about themselves. In contrast, libraries have not given users the ability to make choices about their personal data. Instead, we assume that all users want all their personal data protected at all costs. In this new age, users can choose which bits of information about themselves they wish to share with the world.

Librarians need to explore other possible avenues for gaining information about user preferences. We could ask for usernames in tools like del.icio.us, LibraryThing, and CiteULike and work with the materials the user has added to these services to make recommendations.

There are two distinct advantages to this model. First, the library is not “collecting” information about its users and as a result doesn't have the information safeguarded per se. Second, users can work with existing tools to designate their preferences.

Libraries should allow users to make the same choices regarding library information as they do about other personal preferences and data they are making available in other social environments on web. This will require libraries to change the nature of the way in which they interact with users. We will need to trust our users' ability to set their own levels of privacy and information sharing. This is exactly the step that Ann Arbor District Library, MI, has taken with library user information. By default, a user's borrowing history is purged.

However, users can choose to have their borrowing history retained and employed by the library to provide recommendations of relevant materials similar to Amazon's “Customers who bought this item also bought” feature. Additionally, the Penn tags project at the University of Pennsylvania has allowed users to contribute information about library materials.

These projects are only beginning steps. Web 2.0 has given libraries the opportunity to redefine their relationships with users by assisting them in establishing the social context of materials. In a world where users are experiencing information overload, such services are extremely valuable. Yet, to reach this goal, libraries must overcome the challenge of providing personalized services without turning our backs on issues of user privacy.