Building intelligent systems out of computers has been a continuous challenge for many computer scientists and developers. Among different paths to that goal, one that has been largely studied involves the explicit representation of knowledge, and the processing of those representations by generic reasoning engines. The advent of the Web, and then of mobile computing, has however dramatically changed the way we use computers, and with it our expectations of what such intelligent systems should be. It has also changed the means available to build them. The goal of this dissertation is to show how, in my work in the last ten years, I have been aiming at novel approaches to knowledge engineering, intending to tackle the new challenges and opportunities brought by the Web.Knowledge-based AI has mostly developed on the premise that knowledge was rare, and as such should be made as stable as possible. A large part of our work has been trying to leverage the problems faced by any knowledge-based system when its context changes. Indeed, is not adaptability a core aspect of intelligence? But adaptive reasoning mechanisms must take into account, from the ground up, the dynamics of their knowledge base. It requires to embrace the fact that information is inherently ambivalent, that it acquires meaning (and hence becomes knowledge) only in the context of a particular problem or task. We have been pursuing a user-centered approach, where data collection and reasoning processes are as transparent as possible, and where meaning is not a pre-defined property of information, but negotiated and co-constructed with users.I first present the theoretical framework that we have proposed to build knowledge-based systems exploiting activity traces. By capturing the inherent complexitiy of the user’s task, this kind of knowledge allows for multiple interpretations, and hence requires a special kind of reasoning as well. Then I present a number of our works focusing on assisting a users in her task. One way is to simply present her with her traces in order to help her remembering them and sharing them with others. Another way is to use traces to detect failures and errors, and make helpful proposals for completing the task. The next chapter describes our activity related to Web technologies and Web standards. I show how the foundations of the Web accommodates and even encourages ambivalent information. As such, it allows to bridge the gap between documents, data and knowledge representations. In the next chapter, I focus on a specific class of Web documents, namely hypervideos. I present the models and tools we have proposed to process hypervideos, centered on the notion of annotation, and flexible enough to allow the emergence of new usages.Finally, in the last chapter, to synthesize all the presented works, I propose the groundwork of a theoretical framework for knowledge representation, aimed to cope with, and account for multiple interpretations. In other words, it is an attempt to formalize ambivalent information and the dynamic reasoning processes that use them, allowing to build systems to adapt to the users, rather than forcing the users to adapt to them.