The former PhD student at the IMDEA Software Institute, Isabel García-Contreras, wins one of the two best thesis awards of SISTEDES (Software Engineering and Developing Technologies Society). This is the second recognition she receives after being awarded one of UPM’s best PhD thesis for the 2020/2021 academic year.
“A scalable static analysis framework for reliable program development exploiting incrementality and modularity” is the title of his thesis, directed by professors Manuel Hermenegildo and José Francisco Morales.
Isabel studies scalable analyses in the context of abstract interpretation. Since a way to improve scalability is to perform coarser abstractions, she first inspects what effect this may have in effectively proving the absence of bugs. Second, she presents a framework for scalable static analyses which is generic, that is, independent of the data abstraction of the program. Isabel presents several algorithms for incrementally reanalyzing whole programs in a context-sensitive manner, reusing as much as possible previous analysis results. A key novel aspect of the approach is to take advantage of the modular structure of programs, typically as defined by the programmer, while keeping a fine-grained relation between the analysis result and the source program.
Additionally, she presents a mechanism for the programmer to help the analyzer in terms of precision and performance by means of assertions. She shows that these assertions together with incremental analysis are especially useful when analyzing generic code. All these algorithms have been implemented and evaluated for different abstract domains within the CiaoPP framework.