Camilo
Munoz,
MMath
candidate
David
R.
Cheriton
School
of
Computer
Science
Thanks to the advance in mobile and touch screen devices, handwritten input has gained more popularity among users. When considering mathematical input, however, handwritten math interfaces have to deal with new problems and issues not found in natural language. A popular area of interest that deals with math formulae recognition is math information retrieval (MIR).
MIR starts with a user query and produces a set of results or matches for the query. Applying the concept of pattern matching we can create math formula query languages, which take advantage of the use of wildcards to allow for more expressiveness in the creation of queries. For example, the wildcard # (number sign) could represent a matrix. Unfortunately, the integration of formula query languages and pen-based interfaces is scarce. Therefore, we propose novel query languages suitable for pen-based interfaces and implement a proof of concept using MathBrush, an existing pen-based system.
We propose multiple interfaces that make use of our languages. Lastly, we present an experimental framework that can be used in future research.