Abhishek Shrivastav Avishek Ghosh Jagadish M Sagar Bijwe

Abhishek Shrivastav Avishek Ghosh Jagadish M Sagar Bijwe Blackboard System Applications Guided By Dr. Pushpak Bhattacharya

Collaborative Intelligence

Blackboard Metaphor Imagine a group of human specialists seated next to a large blackboard. The specialists are working cooperatively to solve a problem, using the blackboard as the workplace for developing the solution. Problem solving begins when the problem and initial data are written onto the blackboard. The specialists watch the blackboard, looking for an opportunity to apply their expertise to the developing solution. When a specialist finds sufficient information to make a contribution, she records the contribution on the blackboard, hopefully enabling other specialists to apply their expertise. This process of adding contributions to the blackboard continues until the problem has been solved.

Intuitive View • Knowledge sources • A common place to share knowledge • Control Mechanism

Architecture

Blackboard • Stores Solution Objects • Highly Structured • Multiple Hierarchies • Also stores Control Structures

Characteristics • I think, therefore I am • I don't think like you do • If you can draw it, I can use it • What'd you say? • You could look it up • Is anybody there? • It's my turn • Step by step, inch by inch. .

Problem Solving Steps • Initiation by Knowledge Sources • Control decides what to focus attention on • Knowledge Sources modify solutions • Terminated when either solution is found or no further action is possible

Hearsay II A Speech Understanding System

Sources of Errors • The first source of error is inherent in the speaker and its environment. • • • Semantic – Conceptualization Syntactic – Grammar Lexical - Word choice Phonemic – Articulation Other includes, pauses, extraneous sounds or unnessary phrase repeatitions. • The second source of error in the understanding process is intrinsic in the listener.

Broad Overview of Hearsay II • Hearsay II has a loosely hierarchial structure. • Hypothesis - A partial interpretation actually constructed. • Many competing hypothesis may occur at the same time. • Credibility – The likelihood of a hypothesis being correct. • The goal of the problem solver is to construct the most credible overall interpretation. • The fundamental operation in the construction are hypothesis generation, hypothesis combination and hypothesis evaluation.

Hearsay II Problem Solving Model • Hypothesis generation, combination and evaluation are done by Knowledge Sources. • KSs are condition-action pairs. • KSs communicate through the blackboard. • Role of blackboard: • Repository of hypothesis • Communicates messages from one KS that activate other KSs. • Each hypothesis resides on the blackboard. • The only problem - Combinatorial explosion of the KS invocation that might occur

Levels and Knowledge Sources in Hearsay II

Hearsay II Architecture • Scheduler executes the highest priority activity in the scheduling queues. • The activity may be a KS condition or a KS action • Blackboard Monitor keeps track of the changes occuring in the blackboard.

What Hearsay II gave the world. . . • A general framework based on which many new world applications have been built. • Hearsay II was the first blackboard system developed. • Accuracy of Hearsay II is 90%

Einstein says… “Hang in there, we’re half-way done!”

RADARSAT • Canada's first commercial Earth observation satellite • Provides images of Earth for • Ice and ocean monitoring • Arctic surveillance • Detecting ocean oil slicks. • Oceanography • Geology and Agriculture many more

Mission Management Office • Translation of request to detailed schedule of activities. • Automated planning and scheduling system. • Many possible scheduling combinations. • Model each of the system elements to check violation • Physical constraint • Policy constraint

Hurdles in Development of MMO -PLAN component • No experience of gathering requirement of system. • Planning and Coordination for time segments as small as fractions of second. • Sponsorship by number of participating partners. • 140 system constraints. • Changes during the operational life of the system. • No single optimum solution for every combination. • Number of users and user requests • Less amount of time to develop

Motivation for Blackboard System • Extensible System • Incremental planning, flexibility • “automable” system • Reduction in the scope of initial development

Working of PLAN § Policy Constraints § System Constraints Planned Actions KS 1 Constraint 1 KS 2 Constraint 2 KSn Constraint n Violation of constraints =>Removal of planned Action

Result “The system works very well”, explained Mohamad Farhat, Software Systems Manager for the mission. “It pulls in requests from around the world, 24 hours a day, and we've never missed a product [delivery]. ” “ Each time we bring on a new station, we just add a new set of controllers to interface with the receiving station for the download of image data. ”, according to Ken Lord, Project Manager for the operational phase of the mission.

What we learned from PLAN “The most successful applications are never completed they evolve with the enterprises they serve. ” -Blackboard System Provides the Platform.

Benefits of blackboard architectures • Flexibility of configuration • Flexible problem solving • Selection of knowledge sources • Multiple problem solvers • Management of multiple levels of abstraction • Opportunistic cooperation

Why So Few Blackboard Applications? • Do not scale down to simple problems. • Some systems can be reimplemented without the blackboard structure. • Lack of commercial software • Myth : Blackboard applications are too slow or too hard to develop. • A shortage of application developers with experience building BB applications.

Want to develop Blackboard Systems? • Use university research software • GBB • Use in-house expertise • Commercial & open source tools • Open. BBS • Giga. Spaces

Emerging Directions • Parallel Blackboard Architectures • The concurrent execution of knowledge sources (KSs) and control components in shared address space • Distributed Blackboard Architectures • Communication of blackboard data among autonomous blackboard subsystems • Real-time problem solving • Deal with hard real-time constraints imposed by the environment and they must interact with larger systems within which they are embedded

Conclusion More than quarter-century of blackboard-system experience have produced a strong baseline of collaborating-software technologies. Yet, much more research remains in developing high performance, generic collaborating-software capabilities. Further advances in meeting the collaboration challenges (like representation, interaction, integration & coordination) are needed to enable the next generation of complex collaborative-software applications.

References • Lee D. Erman, Frederick Hayes-Roth, Victor R. Lesser, and D. Raj Reddy. The Hearsay-II Speech-Understanding System: Integrating Knowledge to Resolve Uncertainty, Computing Surveys, 12(2): 213 -253, June 1980. • Daniel D. Corkill. Countdown to Success: Dynamic objects, GBB, and RADARSAT-1, Communications of the ACM, 40(5): 48 -58, May 1997 • Daniel D. Corkill. Blackboard Systems. AI Expert, 6(9): 40 -47, September, 1991. • Dr. John Hunt. Jay. Dee Technology Ltd. Blackboard Architectures. Date: 27. 05. 2002. Version 1. 0. • http: //www. space. gc. ca/asc/eng/satellites/radarsat 1/