GC 2016 - Fall Symposium: Accelerating Science: A Grand Challenge for AI

As a recent Computing Community Consortium white paper noted, scientific progress in many disciplines is increasingly enabled by our ability to examine natural phenomena through the computational lens, and our ability to acquire, share, integrate and analyze disparate types of data. The emergence of "big data", instead of making the scientific method obsolete as some have suggested, underscores challenges in the development of algorithmic or information processing abstractions of various aspects of the scientific methods and processes; the development of cognitive tools that complement and extend human intellect, in the form of computational artifacts (representations, processes, protocols, workflows, software) to partner with humans on all aspects of science (e.g., mapping the state of knowledge in a discipline and identifying gaps, formulating and prioritizing questions; designing, prioritizing and executing experiments; drawing inferences and constructing explanations and hypotheses from the literature, databases, knowledge bases, expressing and reasoning with scientific arguments of variable certainty and provenance; synthesizing findings from disparate observational and experimental studies; formulating new questions, in a closed-loop fashion); Integration of the resulting cognitive tools into collaborative human-machine systems and infrastructure to advance science, including tools for documentation, replication and communication of scientific studies, collaboration, team formation (incentivizing participants, decomposing tasks, combining results, engaging participants with different levels of expertise and abilities), communication (across disciplinary boundaries and across levels of abstraction), tracking scientific progress and impact.
The AAAI Fall Symposium on Accelerating Science: A Grand Challenge for AI (co-sponsored by AAAI and the CRA Computing Community Consortium) aims to bring together researchers in relevant areas of artificial intelligence (e.g., machine learning, causal inference, knowledge representation and inference, planning, decision making, human computer interaction, distributed problem solving, natural language processing, multi-agent systems, semantic web, information integration, scientific workflows), high performance data and computing infrastructures and services, and selected application areas (e.g., life sciences, learning sciences, health sciences, social sciences, food energy and water nexus) to discuss progress on, and articulate a research agenda aimed at addressing, the AI grand challenge of accelerating science.
Topics
Representative topics of particular interest include, but are not limited to:
Computational abstractions of scientific domains, including: Representations of entities, properties, relations, processes of interest in specific scientific disciplines and formal methods for their analysis and simulation; Formalisms for specification of models that take into account uncertainty, and variability;
Methods and tools for linking models across multiple levels of abstraction and spatial and temporal granularity;
Cognitive tools that augment and extend human intellect, including methods and tools for: Charting the current state of knowledge in a discipline and identifying the major gaps; Formulating and prioritizing questions that are ripe for investigation; Machine reading, including methods for extracting and organizing descriptions of experimental protocols, scientific claims, supporting assumptions, and validating scientific claims from scientific literature, and increasingly scientific databases and knowledge bases; Literature-based discovery, including methods for drawing inferences and generating hypotheses from existing knowledge in the literature (augmented with discipline-specific databases and knowledge bases of varying quality when appropriate), and ranking the resulting hypotheses; Scientific Argumentation, including expression, reasoning with, updating scientific arguments (along with supporting assumptions, facts, observations), including languages and inference techniques for managing multiple, often conflicting arguments, assessing the plausibility of arguments, their uncertainty and provenance; Observation and experimentation, including languages and formalisms for describing and harmonizing the measurement process and data models, capturing and managing data provenance, describing, quantifying the utility, cost, and feasibility of experiments, comparing alternative experiments, and choosing optimal experiments (in a given context); Exploration, including navigating the spaces of hypotheses, conjectures, theories, and the supporting observations and experiments; Analysis and interpretation of experimental and observational data, including machine learning methods that: explicitly model the measurement process, including its bias, noise, resolution; incorporate constraints e.g., those derived from physics, into data-driven inference; close the gap between model builders and model users by producing models that are expressible in representations familiar to the disciplinary scientists; Synthesis of findings in a target setting from disparate experimental and observational studies (e.g., implications to human health of experiments with mouse models); Reproducible science, through documentation, sharing, review, replication, and communication of entire scientific studies in the form of reproducible and extensible scientific workflows; Integration of scientific findings into the larger body of knowledge within or across disciplines; Collaborative science, including representations, processes and tools for collabortation, communication, incentivizing participants, and forming teams with with complementary knowledge, skills, expertise to address complex problems (including those that span disciplinary boundaries or levels of abstraction); Citizen science, including tools for decomposing tasks, assigning tasks, integrating results, incentivizing participants, and engaging large numbers of participants with varying levels of expertise and ability in the scientific process; Science of science, including tracking scientific progress, the evolution of scientific disciplines and scientific impact;
Multi-disciplinary, interdisciplinary, and trans-disciplinary applications of AI to accelerate science across a variety of disciplines (life science, learning sciences, health sciences, social sciences, urban sciences, sustainability) that bring together: Experimental scientists in a discipline, e.g., the biomedical sciences, with information and computer scientists, mathematicians, etc., to develop algorithmic or information processing abstractions to support theoretical and experimental investigations; Organizational and social scientists and cognitive scientists to study such teams, learn how best to organize and incentivize such teams and develop a science of team science; Experimental scientists in one or more disciplines, computer and information scientists and engineers, organizational and social scientists, cognitive scientists, and philosophers of science to design, implement, and study end-to-end systems that flexibly integrate the relevant cognitive tools into complex scientific workflows to solve broad classes of problems in specific domains, e.g., understanding complex interactions between food, energy, water, environment, and populations.
Symposium Format
The symposium will consist of: an opening session to introduce the symposium topics, goals, participants, and expected outcomes, and several sessions consisting of invited as well as contributed talks and panels, breakout sessions, and a concluding plenary session summarizing the symposium.
Submissions
Interested participants are invited to contribute 2-5 page extended abstracts or position papers that summarize research challenges and opportunities or recent progress in the relevant areas of AI, computer science, or their applications in accelerating science. Authors may be invited to publish substantially revised and extended versions of their abstracts in a special issue of a journal or an edited collection.
Extended abstracts or position papers should be submitted at:
https://easychair.org/conferences/?conf=asai2016
Key Dates
August 5: Submission of contributed abstracts/position papers
August 12: Author notifications
September 8: Camera-ready submissions of accepted contributions
September 23: Invited participants registration deadline
October 21: Final (open) registration deadline
October 21: Symposium schedule finalized
November 17-19: Symposium, Westin Arlington Gateway in Arlington, Virginia
Organizers
Vasant G. Honavar (Symposium Chair), Pennsylvania State University
Carla Gomes (Symposium Co-Chair), Cornell University
Chitta Baral, Arizona State University
Ann Drobnis, Computing Community Consortium
Gregory D Hager, Johns Hopkins University
Sponsors: AAAI, Computing Community Consortium, North East Big Data Hub