Isaac, Marina Jennifer (2020) Incremental parsing algorithms for speech-editing mathematics and computer code. (PhD thesis), Kingston University, .
Abstract
The provision of speech control for editing plain language text has existed for a long time, but does not extend to structured content such as mathematics. The requirements of a user interface for a spoken mathematics editor are explored through the lens of an intuitive natural user interface (NUI) for speech control, the desired properties of which are based on a combination of existing literature on NUIs and intuitive user interfaces. An important aspect of an intuitive NUI is timely update of display of the content in response to editing actions. This is not feasible using batch parsing alone, and this issue will be more serious for larger documents such as computer program code. The solution is an incremental parser designed to work with operator precedence (OP) grammars. The contribution to knowledge provided by this thesis is to improve the efficiency in terms of processing time, of the OP incremental parsing algorithm developed by Heeman, and extend it to handle the distfix (mixfix) operators described by Attanayake to model brackets and mathematical functions. This is implemented successfully for the TalkMaths system and shows a greatly reduced response time compared with using batch scanning and parsing alone. The author is not aware of any other incremental OP parser that handles such operators. Furthermore, a proposal is made for modifications to the data structures produced by Attanayake's parser, along with appropriate adjustments to the incremental parser, that will in the future, facilitate application of OP grammar to program code or other structured content by changing the definition of its content language.
Actions (Repository Editors)
 |
Item Control Page |