| 2004 Technology
Symposium > Intelligent Information Processing
Intelligent Information Processing
Intelligent Information Processing investigates technologies, tools,
and processes that support the discovery, processing, exploitation and
dissemination of information, tools and knowledge. Intelligent agents
are covered in this area.
Audio Hot Spotting
Qian Hu, Principal Investigator
Location(s): Washington and Bedford
Problems
Large volumes of recordings require rapid retrieval of segments potentially relevant to a given query (audio hot spotting). Because of high automatic speech recognition (ASR) word error rates and the loss of important audio information in speech transcription, spoken document retrieval systems that simply combine ASR with information retrieval (IR) do not meet this need in real applications. Objectives
We propose to research and develop audio-specific retrieval algorithms in critical domains by (1) exploiting multiple types of acoustic information from the audio signals; (2) exploring several adaptive techniques to improve existing ASR performance; and (3) fusing component technologies such as ASR, language/speaker identification, audio feature extraction, and information retrieval. Activities
We will research algorithms and techniques to extend and improve ASR and audio feature extraction and to develop audio-based query algorithms making use of the multiple types of audio information. We will research and develop fusion algorithms to build an audio hot spotting system based on the extended ASR, audio feature extraction, language/speaker identification, and the new audio query language. Impact
Our research in audio hot spotting algorithms and prototype development will address the needs of MITRE sponsors with warehouses of recordings waiting for efficient retrieval. It will extend MITRE's information retrieval capability from text to include audio. The expertise gained through the research will equip MITRE to better advise industry developers and our sponsors on audio information retrieval topics and evaluation standards.
^TOP
Distributed Resource Brokering in Complex Network Environments
Paul Silvey, Principal Investigator
Location(s): Washington and Bedford
Problems
Many challenging problems facing MITRE's sponsors require the coordinated use of large numbers of distributed IT resources. Hard computational problems such as data mining become tractable when thousands of computers are simultaneously brought to bear. Likewise, global information management architectures, like those envisioned for a Joint Battlespace Infosphere, require distributed infrastructures to be sufficiently scalable, fault-tolerant, and timely. Objectives
We are investigating the performance of various proposed techniques for distributed resource discovery in peer-to-peer (P2P) networks by modeling and simulating them in realistic complex network environments. By studying many approaches in many situations, we aim to discover fundamental principles that designers of distributed IT systems can use to achieve desired levels of performance in particular real-world environments. Activities
We are experimenting with strategies for simultaneously increasing efficiency and robustness in P2P networks that use flooding-based protocols for distributed search and indexing. By making the networks self-organizing and self-healing, we can demonstrate mechanisms that allow networks of resource brokers to adapt as their environments change. Our modeling and simulation studies are looking at adaptive network topologies and network joining protocol behaviors. Impact
Our sponsors will be affected by the changes that network-centric, distributed, and P2P technologies are bringing to resource discovery and management problems. Dynamic information management in intelligence, logistics, weather reporting, and command and control all face scalability and timeliness challenges that centralized content indexing approaches will not completely solve. This research is helping to improve our understanding of key issues.
^TOP
Foundations for Next Generation Information Access
Warren Greiff, Principal Investigator
Location(s): Washington and Bedford
Problems
Computerized support for information gathering is fragmented across multiple research communities, and integration is difficult due to the lack of an underlying formalism that cuts across the different technologies. Statistical techniques for individual components have been developed in isolation and without a common theoretical foundation. As a result we are left with a number of reasonably effective, semi-principled, incompatible techniques. Objectives
The principal objective is the development of statistical foundations for information access. A successful foundation will comprise rigorous characterizations of the issues of modeling and estimation, together with principled methodologies for adapting to new languages, genres, information domains, auxiliary knowledge sources, and tasks. Activities
We will develop simulations that model the stochastic generation of latent document features, observable document features, the determination of document relevance, and the distribution of query characteristics. We will perform exploratory data analysis on available research corpora to verify our models. A central focus will be on research into the importance of variance reduction and the potential benefits of various bias-variance strategies. Impact
This research is directly relevant to existing MITRE projects. The results will allow MITRE to develop information access systems incorporating new sources of evidence and to tailor information systems to meet specific military and intelligence needs. MITRE will then be strategically positioned to set the direction of research into, and development of, next-generation information access technology.
^TOP
Robot Swarms: Command, Control and Counter-measures
Alan Christiansen, Principal Investigator
Location(s): Washington
Problems
Most unmanned systems today are not autonomous; they are teleoperated by human supervisors. Some tasks, such as reconnaissance, naturally call for teams of multiple robots. We seek to reduce the number of human supervisors required to accomplish these tasks. To accomplish this reduction of human effort, the robots must possess considerable autonomy in order to enable effective collaboration. Objectives
We will develop mechanisms that support the construction of effective, reliable robot teams. The project is called "Robot Swarms" because we will consider using the simple mechanisms present in collections of social insects. We will also consider possible countermeasures for defeating our teams, and investigate mechanisms for reliable operation in the presence of those countermeasures. Activities
We will participate in the RoboCup/AAAI Search and Rescue Competitions, and we will, in parallel, develop a demonstration of autonomous building-clearing by a team of robots. Impact
In the future, our sponsors (military, intelligence gathering, homeland security) will need to deploy robot systems and have them function together. The research in this project will provide mechanisms by which the robots can work together effectively.
^TOP
Social Network Management for Homeland Security
Raymond D'Amore, Principal Investigator
Location(s): Washington
Problems
Emerging work in social network analysis and complex systems provides formalisms for characterizing the behavior of complex organizations. A key focus is organizational "fitness": the ability to adapt and survive. However, what constitutes an adaptive organization and what would be optimal organizational structure? How can organizations be "assembled" or adapted to respond to changing needs? Can adaptive organizations be "managed"? Objectives
The research will focus on tools for leveraging MITRE's knowledge assets through exploitation and management of enterprise social networks. Social networks as a management tool can transform MITRE internal organizations into "Virtual FFRDCs" and provide a way to better align MITRE resources to sponsor needs. The initial focus will be on MITRE's support to the Department of Homeland Security (DHS). Activities
Initial requirements analysis will identify a range of management needs in the DHS area, which will focus the research and prototype development required to develop tools for such applications as support for ad hoc team formation, technology tracking, and MITREDHS resource alignment. The primary deliverable will be several working prototypes consistent with prioritized requirements and project resources. Impact
Knowledge workers will have new tools for tracking who knows what, and how people and other resources are aligned with mission areas. These tools will facilitate the organizational adaptation needed to respond to emerging requirements. Many of these tools may have applicability in our sponsor environments and may provide a basis for more effective cross-boundary information flow and collaboration.
^TOP
|
