Citation Context

... metrics, complexity metrics(including information flow metrics), understandability metrics, object-oriented metrics and so on 1 . Of particular interest to my problem scenario are code reuse metrics(=-=Kontogiannis, 1997-=-), as JavaScript instances are often copied and modified from standard examples. In my experiments, I assess the predictive strength of these metrics on program categories. I believe that a standard t...

Citation Context

...t and unrelated. To connect these forms together, I need to normalize these non-standard words (NSW) to resolve this feature mismatch problem (Sproat, Black, Chen, Kumar, Ostendorf, & Richards, 2001; =-=Rowe & Laitinen, 1995-=-). Example Transition Patter Split Result onMouseOver single lowercase ⇔ single uppercase on Mouse Over IPAddress consecutive uppercase ⇒ lowercase IP Address IP Address thisweek not an English word, ...

Citation Context

...y subfields, in which formal methods are favored over approximation. The subfield of program comprehension develops models to explain how human software developers learn and comprehend existing code (=-=Mathias et al. 1999-=-). These models show that developers use both top-down and bottom-up models instheir learning process (von Mayrhauser & Vans 1994). Topdown models imply that developers may use a model of program exec...

Citation Context

...els in their learning process(von Mayrhauser & Vans, 1994). Top-down models 4simply that developers may use a model of program execution to understand a program. Formal analysis via dynamic analysis (=-=Blazy & Facon, 1998-=-; Gschwind, Oberleitner, & Pinzger, 2003) may yield useful evidence for categorization. Program comprehension also employs software metrics, which are units of statistical measurements of certain soft...