We are happy to have two exceptional invited speakers at our workshop: Jean-Marie Favre and Michael Tanenblatt.
Invited Speaker: Jean-Marie Favre
Software Anthropologist
Principal Scientist at OneTree Technologies
Luxembourg
Software Linguistics and Software Language Engineering as Emerging Disciplines
Informatics is about automatic symbol processing. Without languages no information could be shared. As a result Linguistics, the science of languages, should be recognized as one of the roots of Informatics. This talk concentrates on Software Languages, that is the set of ALL languages that constitute the essence of Software and/or that are used to talk about Software. This includes programming languages, but even more importantly a much wider range of languages such as requirement languages, modeling languages, rule-based languages, domain-specific languages, markup languages, etc. What is more, Sofware Languages can take different forms and/or and can be found under different names: metamodels, grammars, ontologies, schemas, interchange formats, APIs, etc. Historically these language-based artefacts have been studied in isolation by separated communities. We claim however in this talk that they should be considered under the common umbrella of Software Language. In this context, Software Linguistics is defined as Linguistics applied to Software Languages, and Software language Engineering as Software Engineering applied to Software Languages. As such Software Language Engineering is a generalization of Grammar Engineering, Ontology Engineering, Metamodel Engineering, etc. The relationships between Software Linguistics, Software Language Engineering and other Language research topics such as Computational Linguistics and Natural Language Processing are briefly sketched. As Software Languages are always hybrid and include both Natural Language and Formal Language components, we advocate for a better integration of Software Engineering and Natural Language Processing.
Bio
Jean-Marie Favre is Software Language Archeologist and Software Anthropologist. He got a PhD in Computer Science at the University of Grenoble, France. He is a principal scientist at One Tree Technologies (Luxembourg) and serves as Assistant Professor at the University of Grenoble. His research interests include the evolution of very large software product, reverse engineering, the history of informatics, software languages and research 2.0. He has collaborated with major players in software industry, and he is PC member various Software Engineering conferences and workshops and has organized more than a dozen of international events. He is member of the steering committee of the International Conference on Software Language Engineering. He has co-edited a book on Model Driven Engineering (in French), and various special issues of international journal, including a special issue on Software Language Engineering in IEEE Transaction on Software Engineering. He practices actively Research 2.0 and XFOR 1.0.
Invited Speaker: Michael Tanenblatt
IBM T.J. Watson Research Center
USA
Abstract
One of the benefits of using a framework such as UIMA is that it is easy to assemble and deploy pipelines of analytics components. Unfortunately, the goal of an easy interchange of components with those of similar functionality from alternative sources is usually hindered by mismatches in type systems, as well as other basic assumptions. The goal of building libraries of reusable components for text analysis that can be easily shared has therefore been elusive.
In this talk, I will discuss some of the problems of interoperability and component reuse, as well as some of the different solutions and their drawbacks, hopefully provoking a discussion as to how some of these gaps can be bridged.
