The next day, Bob launches the topics dashboard web application to revise the data he has collected. All extracted data items can be browsed by type, so Bob navigates to the list of movies and opens the details view of a recently added movie. In order to add some details manually and classify the movie, he switches to the edit mode (Figure 10) and decides to add the movie to his list of favourite movies. Subsequently, he categorises the conference events according to their areas of interests (Figure 11). It is important to note that topics are not tied to a specific semantic type, but may be used to classify different types into the same category. This would allow Bob for example to group information about researchers, conference events and conference locations under the same topic. As a result, Bob only needs to bookmark the URL of the topic instead of having to bookmark all the web pages individually. In addition, the extracted and locally stored information provides him with a concise overview of all the relevant information and thus makes visiting the individual web pages obsolete. Missing information can be added in-place and does not need to be maintained separately. By gathering information only from user-selected web pages, our system allows Bob to create an individual, personal data repository with ease.