The continued high growth rate of research publications in bioinformatics has led to a torrent of new data relevant to understanding life and promoting human health. To keep up, the biological community has built up some very large ontologies (many tens of thousands of terms) identifying the basic phenomena of interest, as well as databases of organism models linking these terms to particular genes of interest, based on published evidence.
Carefully managed databases provide structured and continuously updated information from a dizzying variety of biomedical data sources. The global move toward electronic patient records and patient-based social networking sites has the potential to provide far deeper insight into disease progression and the effectiveness of therapy than ever before possible. However, the computational challenges in effectively exploiting all of this information are immense.
In this talk, I will describe some surprisingly effective, recent computer systems that integrate natural language processing, semantic data integration, automated inference, and visual analytics to support knowledge-based data analysis at genomic scale. I will also speculate about future developments, argue that biomedicine provides the most likely context for the first solution of an AI-complete problem, and try to explain why that is.Bio:
Dr. Lawrence Hunter is the Director of the Computational Bioscience Program and of the Center for Computational Pharmacology at the University of Colorado School of Medicine, and a Professor in the departments of Pharmacology and Computer Science (Boulder). He received his Ph.D. in computer science from Yale University in 1989, and then spent more than 10 years at the National Institutes of Health, ending as the Chief of the Molecular Statistics and Bioinformatics Section at the National Cancer Institute. He inaugurated two of the most important academic bioinformatics conferences, ISMB and PSB, and was the founding President of the International Society for Computational Biology. Dr. Hunter's research interests span a wide range of areas, from cognitive science to rational drug design. His primary focus recently has been the integration of natural language processing, knowledge representation and machine learning techniques and their application to interpreting data generated by high throughput molecular biology.
Franz created a first prototype of AllegroGraph in 2005 for a DOD conference on the Semantic Web. It was so well received that we have been working on it ever since. Currently we have more than 15 active, full time developers working on the product and we are at version 4.0. We have a respectable user base that is helping us to fund further development.
In my talk I will cover several topics.
Dr. Jans Aasman, CEO of Franz Inc. started his career as an experimental and cognitive psychologist, earning his PhD in cognitive science with a detailed model of car driver behavior using Lisp and Soar. He has spent most of his professional life in telecommunications research, specializing in intelligent user interfaces and applied artificial intelligence projects. From 1995 to 2004, he was also a part-time professor in the Industrial Design department of the Technical University of Delft. Jans is currently the CEO of Franz Inc., the leading commercial supplier of Common Lisp and scalable RDF database products that provide the storage layer for powerful reasoning and ontology modeling capabilities for Semantic Web applications. Dr. Aasman has gained notoriety as a conference speaker at such events as Semantic Technologies Conference, International Semantic Web Conference, Java One, Linked Data Planet, INSA, GeoWeb, ICSC, RuleML and DEBS.
Gambit is a complete development system for the Scheme programming language which includes an optimizing compiler, and interpreter and debugger. The system has evolved over more than 20 years from a research project into a robust implementation suitable for commercial development. It has been used for implementing other languages, distributed systems, and computer games. The talk presents an overview of the system from a user's perspective and also its evolution and implementation, particularly the portable Scheme to C compiler.Bio:
One of Common Lisp's great strengths is the high-quality ANSI standard. However in these days of popular and successful languages defined by de facto specifications, dominant implementations, and benevolent dictators, not everyone sees the standard as an unmitigated benefit. Especially frustrating to many new Lispers is the seeming impossibility of reopening the standard to just fix a few things or to standardize a few areas not currently covered. This talk will discuss the history of how the Common Lisp standard came to exist, why it will probably never be changed, and what that means for the future of Common Lisp specifically and Lisp generally.Bio:
Peter Seibel is either a writer turned programmer or programmer turned writer. After picking up an undergraduate degree in English and working briefly as a journalist, he was seduced by the web. In the early 90s he hacked Perl for Mother Jones Magazine and Organic Online. He participated in the Java revolution as an early employee at WebLogic and later taught Java programming at UC Berkeley Extension. He is also one of the few second generation Lisp programmers on the planet and was a childhood shareholder in Symbolics, Inc. In 2003 he quit his job as the architect of a Java-based transactional messaging system, planning to hack Lisp for a year. Instead he ended up spending two years writing the Jolt Productivity award-winning Practical Common Lisp. His most recent book is Coders at Work, a collection of Q&A interviews with fifteen notable programmers and computer scientists. When not writing books and programming computers he enjoy practicing Tai Chi. He live in Berkeley, California, with his wife Lily, daughters Amelia and Tabitha, and dog Mahlanie.
The release of F# 2.0 makes functional programming a viable choice for mainstream development in the context of .NET. We'll look at the evolution that set the scene for F#: the web and multi-core changes that have taken place in the industry to make functional programming more relevant, the long and rich history of functional programming itself, through to the technical stuff: the introduction generics in .NET, LINQ in C# and the evolution of F# itself. We'll look at F# today including its parallel and asynchronous programming support, and sneak a preview of F# 3.0 as we integrate a world of data into the functional programming experience.Bio:
Don Syme is a Principal Researcher at Microsoft Research, Cambridge. He is the designer of the F# language, recently released as F# 2.0. He is also the co-designer and co-implementer of generics for the .NET CLR, now used as a fundamental part of the programming models of C# and VB in Mono and Microsoft implementations.
The theme of my talk is why Lisp is a great choice for highly ambitious software products, and has consistently been my choice over a nearing-50-year-long career. I will focus on two such choices. The first was made for G2, the real-time expert system product of my 1986 startup, Gensym Corp. -- a choice that was key to G2's becoming arguably the most successful commercial AI product of that era. It is today a 23-year-old product that continues to run mission-critical industrial applications worldwide. The second choice was made for the two products of my 2010 startup, Expressive Database: EXP DB, a novel database that can hold billions of highly semantically and contextually organized expressions, and EXP READ, an natural-language-to-EXP translator, built upon EXP DB, which aims to achieve near-human-level NLU. Among other unequaled capabilities, Lisp's support for data abstraction through macro definition is critically important to the success of all these ambitious products, and Lisp's model of software that can dynamically change as it continues to run in real time was central to many of G2 customers' application successes. For G2, I will illustrate these points with many anecdotal and sometimes outrageous application examples -- at NASA, at Aughinish Alumina, at Intelsat, at Lafarge, at Siemens, and at Nabisco. For EXP DB and EXP READ, I will explain how various of Lisp's good ideas can be adapted to the AI Holy Grail problem of NLU ("Natural Language Understanding"), how Lisp is needed to support the simple-on-the-surface-but-complicated-underneath data abstractions of expressive databases, and why I believe NLU will first be seriously "solved" using expressive databases (or something akin thereto). I plan to start my talk with a few nostalgic reminiscences from the early days of G2, and end it by reiterating how critically important Lisp can be to the success of highly ambitious software products.Bio:
Lowell Hawkinson has had a long, Lisp-based career, beginning at Yale
as an undergraduate in 1962. There, he built a Lisp interpreter and
compiler called YULISP, which had a rather sophisticated relocating
garbage collector -- perhaps the first instance of this kind of
garbage collector. From 1963-65, he worked under the mentorship of
Harold V. McIntosh while at various university postings (McIntosh
sponsored the first International Lisp conference, in Mexico City in
1963.) From 1965-67 he led the effort to implement Lisp 2, a
John-McCarthy-sponsored and ARPA-funded intended successor to Lisp
1.5. From 1973-83, he worked in Lisp on natural-language-based
knowledge representation at MIT's Laboratory for Computer Science. In
1983 he went to Lisp Machines, Inc., first as a general manager, then
as development manager for PICON, a real-time expert system tool built
in Lisp. From 1986-2007, with a couple of breaks, he was chairman &
CEO of Gensym Corp., a software company with a Lisp-based real-time
expert system product G2 that grew from an 8-person startup to a 300+
person public company. As of 2010 he has a new startup, Expressive
Database, that is using Lisp for building a novel "expressive
database" product, and then an NLU product on top of that.
Copyright © 2010 Association of Lisp Users | contact: