Problems and Possibilities

Email is the de facto Personal Information Manager. Email's role as conduit naturally leads it to be used for three key Personal Information Management (PIM) functions: task management , personal archiving and contact management .

Many people exploit the inbox for task management. They leave information about current tasks in the inbox, knowing that when they access it to pick up new messages, they will be reminded about those tasks. They even send themselves email to create a message in the inbox as a reminder and perhaps a link to useful information. Some users also organize emails that relate to current tasks into active folders, returning to these folders when they need to deal with those tasks. Email is also used for personal archiving and storing contact information. Reference information delivered through email, or information about completed tasks is often filed in email folders for future use.

Two problems arise from PIM functions being performed in email: fragmentation and lack of direct support for PIM functions. Fragmentation occurs because information delivered through email may be left there rather than relocated to dedicated PIM applications, such as a contact manager, calendar, todo list or the user's file system. Related documents are stored partially in email and partially in the file system - making it hard for them to collate information. On the other hand, Email was not originally designed as a PIM application, leading to a lack of direct support for PIM functions.

Two different techniques have been proposed to address these problems: centralization and information extraction.

• Centralization addresses fragmentation by locating all PIM in email, and provides direct PIM support by explicitly building PIM functions into email. This is the approach taken by Microsoft Outlook, which aims to provide task management, contact management and calendaring within a single application.
• Information extraction takes the opposite viewpoint, aiming to migrate PIM functions and information from email into dedicated applications to provide direct PIM support. It addresses fragmentation by making email data straightforwardly accessible to those applications.

Though both techniques offer distinct benefits over the current situation, neither currently offers a compelling solution to email and PIM problems. A combination of both is needed - to provide greater dedicated support for PIM within email itself, as well as improved data extraction from email into other PIM applications.

In the next few years, three areas where we expect to see rapid developments in email are in mobile access, text processing and task management.

Mobility . Mobile phones and PDAs are becoming important platforms for many users, but they raise significant HCI challenges. First new interfaces have to be designed to finesse a lack of display space. This may serve to introduce new designs that break away from current standards. The demands of accessing information on the move may also mean that mobile workers will also expect more assistance from systems in identifying important messages for immediate processing, opening the door to filtering and personalization techniques, as well as interfaces that directly support mobile access.

Text Processing . The last 10 years since the appearance of the web have seen a transformation in techniques for text processing, e.g. novel information retrieval algorithms, information extraction techniques, and text classification methods. So far these have been applied to public corpora such as the web, or digital libraries, but recently researchers and companies have begun to apply them to personal information such as email. While these automatic techniques have great potential for addressing pressing problems with email, e.g. spam detection, it is also clear that there are significant issues associated with their integration into working email interfaces. Early data indicates that users are often mistrustful of such techniques - having very high expectations and being somewhat intolerant of errors. It will significantly test the ingenuity and design skills of HCI researchers if we are to successfully integrate these new techniques into email UIs.

Task Management . A final area where progress is needed is in redesigning email to help people better manage their tasks. While novel text processing algorithms might help users better detect and cluster related messages in their Inboxes, there is also a more significant challenge; to design systems that support users in detecting and processing messages associated with important tasks. Currently users are at the mercy of an undifferentiated Inbox in which all messages have equivalent status. We need to devise systems that help users identify important messages quickly so that they can proactively focus on their important tasks, rather than sifting through multiple messages attempting to determine how each message might relate to their outstanding tasks.

Specifically, new developments in desktop search, machine learning, and text processing will also generate new possibilities and opportunities. The increasing capacity of systems to analyze text content and perform ever more powerful search functions will lead to some profound changes in what it is possible to do with email and PIM. We can expect to see new systems that can:

• Provide organization at the task or people rather than only the message level.
• Anticipate the importance of email and prioritize it accordingly.
• Detect obligations and message urgency.
• Provide new visualizations that allow users to view and organize information from multiple related messages.
• Propose actions based on email and make them easier to initiate.

Comments