Methods and apparatus are provided for speech-based data retrieval from a portable communication device, such as a cell phone with an onboard dataset. The method comprises semantically indexing the dataset with domain related keywords that are intuitive to a user familiar with the dataset domain. When the user speaks a keyword, a speech recognition system within the portable device links the keyword to a semantically related data file in the dataset. A processor in the portable communication device retrieves the selected data and converts it for presentation on a visual display and/or an audio output.