Desktop Search, the search across local storage such as a personal computer, is a common practice among computer users. There has been much activity in Web-related Information Retrieval, but Desktop Search has only recently increased in popularity. As the structure and accessibility of data in a local environment is different to the Web, new algorithmic possibilities arise for Desktop Search.
We apply a connectivity analysis approach to the local environment—a filesystem. We describe how it can be used in parallel with existing tools to provide “more useful” ranked results. Our evaluation reveals that such an approach has promise, and we conclude that exploiting the organization of a filesystem is beneficial for Desktop Search.