Snehasis Mukhopadhyay - US grants

Abstract

IIS-9817572
Palakal, Mathew
Indiana University
$101,604 - 12 mos.

DLI Phase 2: A Distributed Information Filtering System for Digital Libraries

This is the first year funding of a three year continuing award. The proposed research is aimed at designing and developing a distributed intelligent information distribution and filtering system that provides personalized information services to the user while minimizing direct user involvement. The system is intended to traverse the internet to retrieve the most relevant information of interest to the user. Information filtering will be realized using a information agents, and will involve integration of advanced concepts and techniques from the domains of artificial intelligence, information retrieval, and distributed object computing. The agents will contain models of network-based dynamic information resources and will have the capability to learn changing patterns of an individual user's interest.

Four key basic research areas to be addressed are: methods for adapting various knowledge structures associated with an agent
new and robust agent architectures
agent collaboration protocols based on a natural or artificial economic framework
agent-driven information service operations

The explosive growth of biological information sources, available over the Internet, has given rise to both opportunities and challenges for biological and medical researchers. The opportunities they provide are both scientific (e.g., understanding the information encoded in elementary biological structures) as well as technological (e.g., new drug discovery). The challenges, on the other hand, lie in how to efficiently discover, among the vast volume of information, the items that are relevant or interesting to a given researcher. The objective of the proposed research is to investigate related basic research problems and develop a biological information delivery system in a collaborative project between computer scientists, information scientists, and biological researchers. The specific plans include developing methods to make the proposed system pro-active (surveying evolving on-line sources for relevant information), personalized (cognizant of a particular researcher's interests), adaptive (able to react to changes in the information sources as well as user interests or objectives), and capable of integrating multi-format data. The impact of this research is a significant enhancement in the ability of students and researchers in biological sciences to efficiently utilize on-line resources, while generating methods for computerized analysis of biological data and providing computerized support for new scientific discovery.

This project is applying advances in digital library (DL) technologies to the emerging domain of bioinformatics, and developing interaction tools that support learning based on identifying and visualizing associations among key dimensions of bioinformatics resources. The project is making these current resources, which are mainly utilized by expert biologists, available to bioinformatics students. The project addresses issues regarding: the use of a wide variety of formats and representations to store bioinformatics information; the application of DL technologies particularly in the realm of data description and exchange; mapping of metadata associated with bioinformatics information to data description standards compatible with DL technologies; and building clients that take advantage of data dissemination protocols such as the Open Archives Initiative in order to support novel browse, search, and analysis functions based on visualizations. The project team is also integrating DL and Grid computing technologies, and utilizes bioinformatics resources of Indiana University such as the Drosophila Genome Flybase, the IUBio Archive that contains euGenes eukaryote genes data, and the Bionet news archive and related software.

PROPOSAL TITLE: ESE: Spatial Interactive Optimization for Restoration of Upland Storage in Watersheds: Community Participation in the Design of Distributed Practices and Alternatives

PRINCIPAL INVESTIGATOR: Dr. Meghna Babbar-Sebens, Assistant Professor, Department of Earth Sciences, Indiana University Purdue University Indianapolis.

CO-PRINCIPAL INVESTIGATOR: Dr. Snehasis Mukhopadhyay, Associate Professor, Department of Computer and Information Science, Indiana University Purdue University Indianapolis.

COLLABORATOR: Dr. Edna Loehman, Emeritus Professor, Department of Agricultural Economics, Purdue University.

Abstract

The alteration of the natural hydrologic cycle due to human activities -- such as deforestation, artificial agricultural drainage systems, urbanization, and residential development has resulted in loss of multiple ecosystem services (e.g. flood attenuation and water quality control) that were previously provided naturally by various landscape features in river basins and watersheds. Re-naturalization of the hydrologic cycle in degraded watersheds has been proposed to replace lost storage on floodplains with upland storage. At the same time, agronomic practices recommended by USDA's Natural Resources Conservation Service can improve water quality and habitat. This research focuses on the design of a distributed upland storage system, with design involving the selection of sites, scales, structural changes, and agronomic practices for agricultural landscapes in a degraded watershed. We will use the Eagle Creek Watershed (HUC 11 watershed) for development and demonstration of design methods. Because there are a large number of alternative sites, scales, and mitigation methods, and because there are multiple criteria for selection of locations and methods, design is complex. Quantifiable criteria for selection include downstream flood volume, cost of mitigation, loss of habitat, etc. As in any decision problem, there are also unquantifiable criteria such as inconvenience and loss of productivity and control for private land holders. Because upland re-naturalization must occur on private land, there must be voluntary agreement for any mitigation measures. To address both complexity and acceptability to land holders, our research will integrate computational tools (GIS, simulation, optimization algorithms, etc.) for quantified criteria with community participation to address un-quantified criteria. Specific objectives are: 1) Develop a simulation-visualization framework for stakeholders to assess mitigation alternatives under conditions of climate change. 2) Develop and investigate interactive, multi-objective, and stochastic optimization approaches for including single and multiple stakeholders' participation to generate preferred alternatives that reflect non-quantitative and local criteria. 3) Compare efficacy of the interactive approach with non-interactive optimization through stakeholder assessment. Though we demonstrate the usefulness of the approach for Eagle Creek Watershed, IN, it can, however, be applicable to many other areas and problems.

This project attempts to develop better methods for Reinforcement Learning and Approximate Dynamic Programming (RLADP), in order to be able to handle decision tasks with greater complexity both in time and in space. Reinforcement learning systems are systems which can learn to maximize any measure of performance or satisfaction, based on their experience of observing their environment, acting on the environment, and receiving feedback on performance, similar to the pain or pleasure which is used to reinforce animal behavior. Current reinforcement learning methods do not learn fast enough to perform well, when their environment is too complex in space or in time. This project will develop new methods to handle that kind of complexity. The team will also have a collaboration with IBM research, and will try to address a testbed problem involving the management of a fleet of plug-in hybrid cars.

Complexity in time will be handled by use of a multiple model approach, connecting various options or skills by evaluation and updating of the landmark states which mark transitions between different regions of state space. This is similar to previous work on decision blocks and modified Bellman equations previously presented at the PI's workshop on learning and adaptive systems, but otherwise is a unique, new an important direction. Complexity in space is addressed by a multiagent approach, based on a kind of spatial decomposition.

Mutual learning can happen between two humans, a human and machine, or between two machines. The first class is of interest to researchers in the field of social psychology. The importance of human machine interactions is being felt in many situations and most recently in the interaction between the human driven and completely autonomous vehicles. The proposed research deals with machine-machine learning to investigate efficient cooperation between machines, but also reveal the limitations of this cooperation. In particular, the research will attempt to answer questions such as whether two agents, although individually using schemes that will result in the desired behavior, may arrive at wrong conclusion using mutual learning.

While the term mutual learning has been used by other investigators in the past, our objective is to investigate it in a quantitative sense within the framework of mathematical systems theory. The problems proposed for investigation include deterministic optimization in high dimensional spaces, stochastic reinforcement learning in static/stationary environments (learning automata) using both deterministic and stochastic schemes, learning in dynamic environments such as the ones described by Markov Decision Processes, and learning/adaptation by multiple agents in dynamic environments described by deterministic or stochastic difference and differential equations. The results on mutual learning obtained during the proposed project will be widely disseminated at national and international conferences as well as the bi-annual Yale Workshops on Adaptive and Learning Systems.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

Strengthening American Infrastructure (SAI) is an NSF Program seeking to stimulate human-centered fundamental and potentially transformative research that strengthens America’s infrastructure. Effective infrastructure provides a strong foundation for socioeconomic vitality and broad quality of life improvement. Strong, reliable, and effective infrastructure spurs private-sector innovation, grows the economy, creates jobs, makes public-sector service provision more efficient, strengthens communities, promotes equal opportunity, protects the natural environment, enhances national security, and fuels American leadership. To achieve these goals requires expertise from across the science and engineering disciplines. SAI focuses on how knowledge of human reasoning and decision-making, governance, and social and cultural processes enables the building and maintenance of effective infrastructure that improves lives and society and builds on advances in technology and engineering.<br/><br/>Watershed management is a top priority for the nation. The management of watersheds must attend to floods, droughts, water quality, and ecological systems. Such management is shifting from centralized gray infrastructure solutions (dams, reservoirs, water-wastewater treatment plants) to approaches that incorporate decentralized and nature-based infrastructure solutions (wetlands, grassed waterways, stream buffers, rain gardens, green roofs). This SAI planning project examines how watershed communities can influence the design of nature-based infrastructure and how the design of such infrastructure can be inclusive of different stakeholders. The design of nature-based infrastructure is most effective when implemented as an interconnected network of landscape practices at multiple geographic locations across the entire rural and urban terrain. This process requires a deep understanding of the physical and biological laws that govern how water is stored, regulated, and routed in the natural and built environments of the watershed, and of the perceptual and cognitive processes related to watershed stakeholders who influence what type of infrastructure solutions are implemented and where. This planning project develops such understanding to enable a holistic, sustainable, and resilient watershed management approach.<br/><br/>This SAI planning project focuses on mathematical and computational frameworks for characterizing and coping with behavioral uncertainty in participatory design of nature-based infrastructure at watershed-scales. The project investigates the role of diverse perceptual and cognitive processes utilized when people interact with digital visualization interfaces during design of nature-based infrastructure. It also develops novel, computational and interactive participatory design methods based on modeling approaches for coping with behavioral and physical uncertainty in the design of climate-resilient solutions at watershed-scales. The project draws on theory and research on human-centered interactive optimization for watershed systems, perceptual and cognitive psychology, and multi-model approaches for adaptive control. The planning activity includes data collection and workshops to develop a use-inspired research plan to support fast, accurate, and adaptive participatory design in watershed communities.<br/><br/>This award is supported by the Directorate for Social, Behavioral, and Economic (SBE) Sciences, the Directorate for Geosciences, and the Directorate for Mathematical and Physical Sciences.<br/><br/>This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.