BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Computational Optimisation Group - ECPv6.15.11//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Computational Optimisation Group
X-ORIGINAL-URL:http://optimisation.doc.ic.ac.uk
X-WR-CALDESC:Events for Computational Optimisation Group
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:UTC
BEGIN:STANDARD
TZOFFSETFROM:+0000
TZOFFSETTO:+0000
TZNAME:UTC
DTSTART:20130101T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=UTC:20140115T151500
DTEND;TZID=UTC:20140115T151500
DTSTAMP:20260505T000408
CREATED:20170124T102140Z
LAST-MODIFIED:20170124T102140Z
UID:572-1389798900-1389798900@optimisation.doc.ic.ac.uk
SUMMARY:Seminar: Numerical Aggregation of Trust Evidence: Its Analysis and Optimisation
DESCRIPTION:Title: Numerical Aggregation of Trust Evidence: Its Analysis and OptimisationSpeaker: Prof. Michael Huth and Dr. Jim Huan-Pu KuoAffiliation: Department of Computing – Imperial College LondonLocation: Room 217 Huxley BuildingTime: 3:15pm \nAbstract. We have designed a language in which modellers can specify trust and distrust signals that\, in their presence\, generate a numerical score\, and where such scores can be combined with aggregation operators to express risk postures for trust-mediated interactions in IT systems. Signals may stem from heterogenous sources such as geographical information\, reputation\, and threat levels. Aggregated scores then inform decisions by generating conditions that compare scores to threshold values of trustworthiness. We developed a generic approach to analysing such conditions by automatically converting them into code for the Satisfiability Modulo Theory solver Z3 from Microsoft Research. This allows us to automatically analyse\, e.g.\, whether a condition is sensitive to the increase of a trustworthiness threshold by a specified amount. We would now like to understand better whether such analysis questions can be expressed in known models as used in optimisation. For example\, let a condition say that the aggregated trust score has to be above 0.5. Solvers such as Z3 seem to be unable to compute the largest interval containing 0.5 such that all values of that interval could be chosen as trustworthiness threshold without changing the behaviour of the condition. On the other hand\, Z3 is perfect for reflecting logical dependencies or inconsistencies between (dis)trust signals that occur in such conditions and are quantifier-free formulas of first-order logic. A prototype implementation of the tool is available at http://delight.doc.ic.ac.uk:55555 \nAbout the speaker.
URL:http://optimisation.doc.ic.ac.uk/event/seminar-numerical-aggregation-of-trust-evidence-its-analysis-and-optimisation/
END:VEVENT
END:VCALENDAR