Large-scale Interoperation and Composition (LIC)

Gio Wiederhold

Computer Science Department, Stanford University

We are engaged in a number of projects that deal with new paradigms for building, maintaining, and securing large computing systems. These often involve intermediary services at various levels: Mediators (viewgraphs), so that sharing becomes scalable.

This work is based on concepts investigated in the development of information mediation; mediator technology is now becoming commercially available.

We believe that the software paradigm (viewgraphs) is changing, and will critically effect the business of program creation and maintenance. It should also effect what should be taught. To this end we are teaching CS 446, Large Scale Software System Construction and performing research on

  1. Managing ontologies, and developing an Ontology Algebra to assure Scalability in our SKC project and
  2. Supporting an Infrastructure for the dta to Information foodchain, based in XML and RDF, within the DARPA DAML initiative, the Ontoweb project and
  3. Compiling High-level Access Interfaces for Multi-site Software (CHAIMS), as well as
  4. developing Security mediators for Trusted Interoperation of Health Care Information (TIHI) and Manufacturing (SAW).

Preservation

We are engaged in an NSF-sponsored Workshop on Preservation and Archiving of Documents and Software, to be held march 26-27, 1999 in Washington DC.

Technology Transfer

A meta issue for all these efforts is improving the transition of technology into practice, which I hope to accomplish through early engagement and support to companies who can help in prototype development and then serve as demonstration and transfer sites to end users.

An example is the work on mediators for fusion of data for Environmental Restoration at the Idaho National Engineering Labs (INEL-ERIS) ).

A bibliography focusing on recent related material is available.


The actual projects are:


Scalable Knowledge Composition

A proct initiated in 1997 with funding from AFOSR to develop an Algebra over Ontologies. See (SKC) for details.

A Simulation Access Language (SimQL)

Initiation contract from ARPA DSO / NIST

The intent is to demonstrate viability of a language (SimQL) which will permit integration and reuse of simulations within advanced information systems.
Started Spring 1996, completed summer 1997.
We have submitted a proposal for a major effort (InSim), and were approved, but not funded.

Participants:

Gio Wiederhold (PI)
Rushan Jiang < rushan@leland.stanford.edu > (CS MS, graduated)
James Chiu (CSD, MS; CS395 project, completed graduated)

Simulations provide the real support for decision-making, by allowing us to make projections into the future. A true decision support system should be able to combine database information, (obtained via SQL), knowledge (obtained via KQML), and projections from simulations (obtained via SimQL). The simulations themselves can use a variety of techniques and respresentations, as spreadsheets, equation solvers, stepwise modelers. Our intent is to extend with SimQL the database paradigm, where data can come from relational, textual, and semi-structured databases, and the AI world, where knowledge has many representations, as suited for the computational task, to the unexploited world of simulations.

A number of presentations have beeen given, and several papers were published:
Integrating Simulation Into Information Systems (htm), presented at Dagstuhl, Germany May 2002.

Build Real Information Systems! (ppt draft)

Wiederhold, Gio: "Information Systems that Really Support Decision-making"; Journal of Intelligent Information Systems; Vol.14, Kluwer, March 2000, pages 85-94.

Wiederhold, Gio: "Information Systems that Really Support Decision-making"; Journal of Intelligent Information Systems; Vol.14, Kluwer, March 2000, pages 85-94.


Access to and Formalization of Building Regulations

Supported by NSF ITR with initial research at CIFE, Dept. of Civil Engineering

See REGNET, under direction of Professor Kincho Law, CIFE.
Project being initiated, looking for an interested and capable student.

Mediators for Environmental Restoration Data (INEL/ERIS)

Under contract from the Idaho National Engineering Labs (INEL)

Funded by IRD resources through Lockheed-Martin Idaho Technologies Co (LMITCo).
Mediator technology to integrate Environmental Restoration data from heterogeneous sources (ERIS).
Planning phase (1) (1995) completed, prototype research.
Phase 2 (1995-1996) completed, prototype in operation at INEL.
Phase 3, started in March of in 1997, completed 1998.

Participants:

Gio Wiederhold
David Maluf < maluf@db.Stanford.edu >
Priya Panchapegasan < bhujanga@leland.Stanford.edu >

Prior participants:
Joshua Hui < wjhui@cs.stanford.edu >
ISX Inc (Nancy Lehrer, Charles Channell, Stephan Erickson), a company currently coordinating information on DARPA's I3 program, which supports much other work on mediators. (Phases 1 and 2)
INEL, LMITCo (Lynn Dean, Toni Hempstead, Dave Schwieder, others)


Trusted Interoperation of Health Care Information (TIHI).

Subcontract with SRI International, XiaoLei Qian, now Steve Dawson (PI)

Security mediators to permit collaboration among enterprises that cannot to share all resources.
In progress since 1 March 1995.

Participants at Stanford:

Gio Wiederhold < Gio@cs.stanford.edu >
Michel Bilello < michel@db.stanford.edu > (Post Doc, medical student, project leader)
Andrea Chavez < aechavez@leland.stanford.edu > (JD, CS MS student, graduated)
Chris Donahue < dewi@db.stanford.edu > (demo, programmer, completed)
Vastala Sarathy < vsarathy@us.oracle.com > (CS MS, graduated)

more.


Secure Access Wrapper: Securing Databases by Access Mediation (SAW).

Subcontract with SRI International, XiaoLei Qian (PI)

A new proposal (control number 96030520) which will apply the secure mediation techniques being developed in TIHI within the system survivability program.
Startdate: September 1, 1996

Participants at Stanford:

Gio Wiederhold < Gio@cs.stanford.edu > (PI)
Michel Bilello < michel@db.stanford.edu > (Post Doc, medical student, project leader)
Jahnavi Akella < jakella@leland.stanford.edu > (CS MS)
Chris Donahue < dewi@db.stanford.edu > (programmer, completed)
Yan Tan < yant@db.stanford.edu > (IE MS student)

more.


Logo

Compiling High-level Access Interfaces for Multi-site Software (CHAIMS)

Intially funded by DARPA ITO as part of the EDCS program and partially continuing with support from ORD.

The CHAIMS effort is investigationg a very-high level (mega-)programming language for software module composition. Its initial compiler was able to generate a variety of invocation sequences for current and developing standards for software interoperation, with a focus on multi-computer, distributed operation. The language will include the ability to set up module interfaces prior to executions, request performance estimates from modules pnrior to their invocation, schedule modules in parallel, monitor execution of invoked modules, interrupt inadequately performing modules, and provide data and meta-information to customer interface modules [WiederholdWegnerCeri: "Towards Megaprogramming"; CACM, Nov.1992].

CHAIMS supports a paradigm shift which is already occurring: a move from coding as the focus of programming to a focus on composition. This shift is occurring invisibly to many enterprises, since there is no clear boundary in moving from subroutine usage to remote service invocation. There are hence few tools and inadequate education to deal with this change.

See http://www-db.stanford.edu/CHAIMS/ for details.


Adaptive Crisis Response (ACR): Intelligent Assistants for Joint-Force Crisis-Response

Funded as a Multi-displinary University Research Intiative (MURI)

Research under this grant to develop and integrate agent-based technologies into Advanced Planning and Decision Theory.
Startdate: Fall quarter 1996/97, withdrawn 1997..

Participants at Stanford

Prof. Michael Fehling (PI) (EES)
Dr. Gregg Courand (EES)
Prof. Carlo Tomasi (CS)
Gio Wiederhold (CS)
Arthur Keller (Research Scientist CS) < ark@cs.stanford.edu >
Joshua Hui (CS) < wjhui@cs.stanford.edu >

Precise workplans are still being developed, but are likely to involve


Mediation to Implement Feedback in Training (MIFT)

Supported by ARPA ITO

Mediator technology to analyze data obtained during simulated, real, and mixed training exercises according to training scenarios. For details see MIFT research plans and
demonstration plans.
Funding started March 1 1996 and terminated in 1997.

Participants:

Gio Wiederhold
Myriad Software (Ted Linden) < Linden@MyriadSfw.com >
David Maluf
Bhujanga Priya Panchapagesan (graduated)
Joshua Hui (graduated)

This project undertook research in automated data abstraction, based on a formalized model of the customer's need for information. Such an abstraction process will be performed by a knowledge-driven subsystem in a computer network which mediates between customers and data resources. The aproach focuses on the crucial issue of data or information overload, which occurs when the volume of data exceeds what a customer can comprehend. This problem is increasing in importance, since improved communication, larger databases, and effective search methods are now providing more material than people can afford to read or analyze.

The specific application is to training data, and the model represents the learning/training scenario. A scenario is intended to fullfill a number of training objectives. After the scenario is executed (and perhaps even during excution) the feedback can help design better or complementary successor scenarios,

Publications:

  • Maluf David A., Wiederhold Gio, Linden Ted and Panchapagesan, Priya, "Mediation to Implement Feedback in Training," CrossTalk: Journal of Defense Software Engineering Software Technology Support Center, Department of Defense, 1997.
  • Abstract: The problem addressed by Mediation to Implement Feedback in Training (MIFT) is to customize the feedback from training exercises by exploiting knowledge about the training scenario, training objectives, and specific student/teacher needs. We achieve this by inserting an intelligent mediation layer into the information flow from observations collected during training exercises to the display and user interface. Knowledge about training objectives, scenarios, and tasks is maintained in the mediating layer. A designer constraint is that domain experts must be able to extend mediators by adding domain-specific knowledge that supports additional aggregations, abstractions, and views of the results of training exercises.

    The MIFT mediation concept is intended to be integrated with existing military training exercise management tools and reduce the cost of developing and maintaining separate feedback and evaluation tools for every training simulator and every set of customer needs. The MIFT Architecture is designed as a set of independently reusable components which interact with each other through standardized formalisms such as the Knowledge Interchange Format (KIF) and Knowledge Query and Manipulation Language.

    Links:

    Project overview -- Mediation to Implement Feedback in Training -- MIFT Applied to Military Exercise Management -- Architecture -- Demonstration -- Information Flow (postscript).

    Indexing and Ontologies for Databases within Commercenet.

    Supported by ARPA TRP funding to EIT Inc.

    Support for Electronic Commerce on the Internet.
    Funding started 1993, database portion is terminated in 1997.

    Stanford Principals:

    Michael Genesererth
    Richard Fikes
    Doug Engelmore
    Arthur Keller
    Narinder Singh (left)

    Participants in the InfoLab:

    Hector Garcia-Molina (Completed) < hector@db.stanford.edu >
    Gio Wiederhold
    Luis Gravano (Completed) < gravano@db.stanford.edu > (graduated)
    Ashok Advani (Completed) < advani@cs.stanford.edu > (graduated)
    David Maluf, Postdoc < maluf@db.stanford.edu >
    Marcus Chan, CS MS student < ccc@db.stanford.edu > (graduated)

    This project provided the GlOSS indexing mechanism to EIT, and is now working on a demonstration of an Ontology Algebra to resolve differences among disparate definitions in commercial vendor catalogs

    Slides explaining our view of Ontologies.
    and how to deal with them by disjointly via an algebra over their Domains.
    or read "Interoperation, Mediation, and Ontologies"; Proceedings International Symposium on Fifth Generation Computer Systems (FGCS94), Workshop on Heterogeneous Cooperative Knowledge-Bases, Vol.W3, pages 33-48, ICOT, Tokyo, Japan, Dec. 1994.

    The concepts for this project are now the subject of research in See Scalable Knowledge Composition (SKC) for details.


    Intelligent Software Agents that Access PDQ and CANCERLIT

    NCI Subcontract from Lexical Technology Inc, Alameda CA

    Funding is by the National Cancer Institute (NCI) via an Small Business (SBIR) grant.
    General PI: Mark Tuttle (tuttle@Lxt.com) Lexical technology.
    Stanford PI: Gio Wiederhold (gio@cs.stanford.edu)
    Stanford administrator: Rosalind Ravasio (ravasio@smi.stanford.edu)
    Funding started 1 Oct.1996 and is administered by the Section on Medical Informatics, and is to extend over two years.

    Stanford Principals and Participants:

    Bob Carlson (Medicine/Oncology)
    Larry Fagan (SMI)
    Gloria Linder (librarian consulatant)
    Wanda Pratt (SMI)

    The project is to provide mediation agent technology, based on the NLM UMLS ontology, for NCI's public knowledge bases. in particular for the NCI CancerNet. Security mediators will also be a concern.

    Stanford Computer Industry Project (SCIP)

    Supported through GSB by the Sloan Foundation

    Research in (OBSCIP) investigates factors of success and failure in the Software Industry. My particular interest are the changes occuring in program creation and maintenance, and their effect on industrial practice and educational requirements.
    Started formally 1 January 1996.

    Participants:

    Prof. William Miller, Avron Barr, Shirley Tessler, other GSB faculty, and Gio Wiederhold (CS) < Gio@cs.stanford.edu >
    Wayne Lim < wlim@reuse.com > (MBA, IE student)
    There are potentially positions for students who are interested in analysis of the business aspects of the paradigm change.

    Payment Methods for Web Access to the ACM Literature

    unsupported

    Participant

    Wei Chen, Starfish Software (Completed)
    See ACM payment prototype description.

    Image databases

    Support

    This work was carried out partially under University support, partially within courses as CS545I, and partially during external internships at IBM, SRI international and NEC Corporation. Some student support was provided by the Stanford University Libraries.

    Participants

    James Z. Wang < wangz@CS.Stanford.EDU >
    Jia Li.
    Oscar Firschein.

    On specific projects
    Chen Li <chenli@CS.Stanford.EDU>
    Peter Mork <pmork@stanford.edu>
    Edward Chang <echang@DB.Stanford.EDU>
    Desmond Chan
    others

    Abstract

    There is a range of projects that fall into this category.

    Publications

    1. James Ze Wang, Jia Li, Gio Wiederhold, and Oscar Firschein: "System for Screening Objectionable Images"; Computer Communications Journal, Elsevier Science, 1998. (abstract in HTML with link to full paper in PostScript)
    2. James Ze Wang, Jia Li, Gio Wiederhold, and Oscar Firschein: "System for Classifying Objectionable Websites";, 1998. (abstract in HTML with link to full paper in PostScript)

    Related work

    Trusted Interoperation of Health care Information.
    Trusted Image Dissemination.