BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Software Defect Mining with Formal Concept Analysis - Christian Li
 ndig\, Dagstuhl\, Germany
DTSTART:20071114T141500Z
DTEND:20071114T151500Z
UID:TALK8146@talks.cam.ac.uk
CONTACT:Timothy G. Griffin
DESCRIPTION:\nIdentifying a defect in a software implementation generally 
 requires its\nspecification. However\, we can find promising "defect candi
 dates" by\nstatistical analysis of the implementation alone. The idea is t
 hat every\nlarge program contains redundancies which may be considered as 
 patterns:\n"a function that calls lock() also calls unlock()".  Any deviat
 ion from\na pattern with many instances is a potential defect.  Unlike man
 y\nprevious methods\, this does not rely on pre-defined patterns or names\
 nbut works purely structural.  The theory of Formal Concepts provides an\n
 elegant and practical framework to describe and compute patterns and\nthei
 r violations. Formal Concept Analysis is related to frequent-item\nmining\
 , which was used previously but lacks the rich algebraic structure\nof For
 mal Concepts. Nonetheless\, the talk is light on math and builds on\nintui
 tion and real-world examples. It demonstrates how to find bugs in C\nprogr
 ams with a simple control-flow insensitive analysis and object\nusage anom
 alies in large Java programs using a flow-sensitive analysis.\n\nChristian
  Lindig is a member of the scientific staff of Schloss\nDagstuhl\, a leadi
 ng international conference center for computer\nscience.  He received his
  PhD from Technical University of Braunschweig\nin 1999 under supervision 
 of Gregor Snelting for his work on formal\nconcept analysis in software en
 gineering. In 2000 he joined Norman\nRamsey's group at Harvard\, co-design
 ed and implemented the initial\nversion of the Quick C-- compiler and work
 ed on the definition\,\nimplementation\, and test of compiler calling conv
 entions. After\nreturning to Germany he joined Andreas Zeller's group in 2
 003 at\nSaarland University and worked on aspect mining\, defect localizat
 ion\,\nand code theft detection. His work on code theft recently received\
 nattention by national and international mainstream media\, including\nDeu
 tschlandfunk and Technology Review.\n
LOCATION:Lecture Theatre 1\, Computer Laboratory
END:VEVENT
END:VCALENDAR