Sign In
 
 
 

Courses 2003

Sort by AttachmentsUse SHIFT+ENTER to open the menu (new window).
SyllabusFilter
Title Some computational aspects of geoinformatics
Teacher Prof. Mike Worboys, Department of Computer Science, Keele University, England
Period Ottobre
Syllabus
Much of the early technology for spatial data handling (e.g. spatial indices, object-oriented spatial data models) is now well understood.
Current preoccupations of computer scientists working with geospatial data will be the subject of these lectures. Topics will include spatial reasoning, qualitative spatial data handling, multi-contextual spatial data models and spatio-temporal information systems.
1. Spatial reasoning will be approached from a formal perspective, looking at developments of Egenhofer's intersection method and the RCC calculus.
2. Proper treatment of qualitative data and integration of qualitative with quantitative spatial information is key if useful human computer interfaces with spatial databases are to be developed.
3. Current and future spatial information systems architectures will be multi-contextual, with data arising from many sources and having users with heterogeneous data models and purposes. The talk will develop a framework in which multiple contexts can be developed.
4. Much useful geoinformation has dynamic and historic components, so the lectures will conclude by examining some of the recent work, including that of the European Chorochronos project, on spatio-temporal systems.
 
 
Title Data Mining to Relational Data Mining
Teacher Saso Dzeroski, Department of Intelligent Systems; Jozef Stefan Institute, Ljubljana, Slovenia; Fosca Giannotti, ISTI-CNR, Pisa; Dino Pedreschi, Dipartimento di Informatica, Univ. Pisa
Period February - April
Syllabus
The goal of the course is twold. First, it gives an introduction to the basics of data mining, i.e., the computational methods for the analysis of large and complicated data sets, aimed at extracting knowledge from data. Such methods are based on algorithmic and statistical techniques, as well as database and machine learning techniques. Second, the course introduces to multi-relational data mining, i.e., knowledge discovery from multi-relational data, a line of research rooted in inductive logic programming, at the intersection of machine learning and programming languages.

Title Logica lineare
Teacher Roberto Di Cosmo, Universite' de Paris VI-VII
Period February-April
Syllabus
La Logica Lineare è una logica costruttiva in cui la nozione di "risorsa disponibile" rimpiazza la tradizionale nozione di verità. La Logica Lineare, proposta da Girard, si è rivelata essere non solo uno strumento utile per una migliore comprensione dei sistemi logici preesistenti, ma anche uno strumento fondamentale in vari campi dell'informatica, dalla realizzazione di interpreti efficaci per i linguaggi funzionali, alla migliore comprensione dei modelli di calcolo, fino ad applicazioni in programmazione per vincoli. Lo scopo del corso è di dare una rapida introduzione ai concetti ed alle tecniche fondamentali della Logica Lineare.
 
Title Some Computational Aspects of Geoinformatics
Teacher Mike Worboys, Department of Spatial Information Science and Engineering University of Maine Orono, USA
Period April 2-14
Syllabus
Much of the early technology for spatial data handling (e.g. spatial indices, object-oriented spatial data models) is now well understood. Current preoccupations of computer scientists working with geospatial data will be the subject of these lectures. Topics will include spatial reasoning, qualitative spatial data handling, multi-contextual spatial data models and spatio-temporal information systems. 1. Spatial reasoning will be approached from a formal perspective, looking at developments of Egenhofer's intersection method and the RCC calculus. 2. Proper treatment of qualitative data and integration of qualitative with quantitative spatial information is key if useful human computer interfaces with spatial databases are to be developed. 3. Current and future spatial information systems architectures will be multi-contextual, with data arising from many sources and having users with heterogeneous data models and purposes. The talk will develop a framework in which multiple contexts can be developed. 4. Much useful geoinformation has dynamic and historic components, so the lectures will conclude by examining some of the recent work, including that of the European Chorochronos project, on spatio-temporal systems.

Title Program Verification
Teacher José Meseguer, University of Illinois at Urbana-Champaign
Period June
Syllabus
Programs can be either declarative (based on a logical system) or imperative (conventional languages). They can also be either sequential or concurrent. All four combinations are possible. The course will address program verification techniques for programs in this general sense using techniques from equational and rewriting loogic to specify their semantics; and of first-order logic, inductive logic, Hoare logic, and temporal logic, to specify properties. Proving tools supporting mechanized reasoning in all these cases will be used. The logical foundations will also be developed.
 
Title Rigorous Approaches to Information Security Problems
Teacher Joshua D. Guttman, MITRE, Bedford, MA 01730-1420, USA
Period October - November
Syllabus
The purpose of this course is to explain a method my colleagues and I have applied to a wide range of different problems within the general area of information security. The method has three steps. First, one constructs a model abstracting key aspects of the problem domain in terms of a few mathematical ideas. Second, one selects one or a few logical forms of statement, meaningful in the model, as the security goals to be achieved in different circumstances or configurations. Third, a combination of proof techniques and algorithms are needed to determine the security goals achieved in given circumstances, or to construct circumstances in which given security goals are achieved.
We will apply this general method in three main areas. First, we will apply it to security management. Security management is the problem of using simple mechanisms on a large scale to achieve predictable security goals. We will consider security management problems for filtering routers, for gateways executing the IP security goals, and in the operating system context of Security Enhanced Linux.
Second, we will apply our method to cryptographic protocols, emphasizing the strand space formalism. Strand spaces give us a systematic way to discover flaws in existing protocols; to prove that they achieve authentication and confidentiality goals; and to design new protocols to meet specific goals.
Finally, we will examine trust management, which is the use of logical theories by multiple principals to derive access control decisions. We will also explain recent work interrelating trust management theories with cryptographic protocols to ensure that access derivations rely only on recent information.