Personal Information Management (PIM) studies the practice of storing, organizing, and retrieving information by an individual in support of their roles and tasks (Bergman, et al., 2004). One important problem in PIM is information fragmentation (IF) — the condition of having data in different formats, distributed across multiple locations, manipulated by different applications, and residing in a generally disconnected manner (Tungare, 2007). IF can conflict with the PIM ideal that users should have access to the right information at the right time, in the right place, in the right form, and of sufficient completeness and quality to perform the task at hand (Bergman, et al., 2004). It is typically assumed that IF is unintentional, and occurs as a result of the many applications and devices we use to do our daily work. It is further assumed that IF is "bad" or has negative consequences. In this thesis, I study when IF occurs intentionally. Intentional IF (IIF) refers to the fragmentation in PIM that occurs when a person fragments his or her own personal data purposefully. Although research into the problem of IF has been growing quickly in the past decade, IIF has not been investigated in the literature. Prior studies have portrayed IF as a problematic type of information management. Email is a common context in which IF is found. While IF in email may be unintentional, such as when required by an employer, it is also likely to be intentional, as is the case when users use separate email accounts for different purposes. To further the research in this field, this project investigated the phenomenon of IIF in email by conducting and analyzing data from an online survey. In addition to finding the extent of IIF in email, the survey addressed what motivates the participant to purposely fragment their email as well as the advantages and disadvantages in doing so. My study is the first that has explored intentional fragmentation of information. The findings of this study show that IIF exists in email usage, revealing that IIF occurs across a user's devices and also across a user's multiple email accounts. The two most common motivations for IIF are to keep information separated by the user's social roles (work, school, personal communications, etc.), and to filter out extraneous information in order to simplify their information management. These results show that in addition to the negative consequences of IF there also exists positive uses of IF, that is helpful for some users. / Master of Science

The web browser has emerged as a central workspace for information workers, where they make use of cloud-based applications to access their information. While this solution nicely supports access to their data from multiple devices, it presents a nightmare for organizing and coordinating data between tools for a single project. Information is typically scattered between various online tools, where storage and organization structures are replicated. Information workers are interrupted and have to switch between projects frequently. Once interrupted, resuming work on a project can be hard. To address this information fragmentation and the impact of work interruptions, I created Contextinator, a personal information manager for the web browser that lets information workers organize their work activity and information into projects. Contextinator assists in coordinating information for projects, thereby ameliorating information fragmentation for projects that live on the cloud. It assists information workers in context switching and resuming work after interruptions. In my the- sis, I describe the problem of information fragmentation in the cloud. I discuss the different areas of related work of Personal Information Management, the design of Contextinator and how it is grounded in previous research. I briefly discuss how Contextinator is implemented. I then present the results from my field-evaluation of Contextinator. Finally, I conclude by discussing future work in this research. / Master of Science

In personal task management, individual tasks are susceptible to failure at various stages due to failures in prospective memory, information fragmentation, and/or cognitive overload. To alleviate the troubles that occur in personal task management, people use various tools such as notifications and external memory aids, but there is still room for improvement in regards to maintaining distributed cognitive artifacts such as calendars and to-do lists. Improvement in staying abreast of this personal task information as opposed to being alerted of it in individual instances was the topic of exploration in this work. TaskAmbient is an ambient information display system that was designed to address these problems. TaskAmbient supports retention of individual task knowledge as well as retention of task knowledge in a user's different areas of responsibility. With this tool, I observed usage to verify or deny its ability to support prospective memory and combat the problems associated with information fragmentation and cognitive overload in respect to personal task management. In this research, studies were conducted to learn about personal task management practices and how TaskAmbient was used. In conducting these studies, I found that TaskAmbient provided value in retaining task information and staying aware of tasks in a user's various areas of responsibility. TaskAmbient showed the extendibility to other domains of personal information management. / Master of Science