Personalization of information search has a long history in information retrieval, starting in the 1960s with the work of Hans Peter Luhn on the selective dissemination of information. One important challenge that has persisted throughout time is how information about users and their interests is obtained and maintained. In this talk, I will review some approaches to personalization with an emphasis on techniques for acquiring information from users about their interests and preferences. Three major techniques will be discussed: explicit feedback, implicit feedback and mining the user's desktop. Examples of each technique will be presented along with a discussion of problems that impact their effectiveness including those related to the user's information seeking context and task. Problems related to the conceptualization of what is being modeled by such techniques will also be discussed. This talk ends with the identification of new opportunities for understanding users and their interests, and using this to personalize search.