Software process modelling is the activity of formalising the production lifecycle of large software systems. Its aim is to formally describe a software development process, which is then effectively used and possibly enacted by an environment able to support the geographically distributed and coordinated activities involved in the process itself. I show that rule-based languages, especially logic programming languages, are an important technology for the specification, modelling, enactment and coordination of software processes. This is because most routine activities in any development process can be defined by rules. Some initial proposals aimed at simply simulating the software process by a Prolog-like program embedding some development rules. A further step toward the integration of rule-based languages in the software process has been taken using a dynamic knowledge base as project database, and a number of special primitives have been introduced to support process programs. Currently there is a trend toward more complex programming environments, called process-centred development environments. I show how some rule-based coordination language have been used to build an environment of this kind.

Deriving formal specifications from informal requirements is extremely difficult since one has to overcome the conceptual gap between an application domain and the domain of formal specification methods. To reduce this gap we introduce application-specific specification languages, i.e., graphical and textual notations that can be unambiguously mapped to formal specifications in a logic language. We describe a number of realised approaches based on this idea, and evaluate them with respect to their domain specificity vs. generality.

Logic programming is a programming paradigm with potential to contribute to software engineering. This paper is concerned with one dimension of that potential, the impact that experience with developing logic programs can have on software design. We present a logic programming perspective on programming patterns, systematic program development, design for provability, and the paradigm of meta-programming.

The complex process of deriving programs from specifications is often divided into the following three steps: (i) the derivation of formal specifications from the informal ones; (ii) the validation of the formal specifications; and (iii) the derivation of executable programs from the formal specifications.

The definitions above, separated by ten years, represent two very different conceptions of learning. For Simon learning depends on an internal change in representation, and for Kaelbling it is instead measured in terms of an external change in behaviour. Furthermore, Kaelbling's focus on the situatedness of the learning system being embedded in its environment reflects the recent experience gained by much direct experimentation with physical robots.

Two phenomena have probably affected modern data analysts' lives more than anything else. First, the size of real-world data sets is getting increasingly large, especially during the last decade or so. Second, modern computational methods and tools are being developed which add further capability to traditional statistical analysis tools. These two developments have created a new range of problems and challenges for analysts, as well as new opportunities for intelligent systems in data analysis.