[comp.doc.techreports] tr-input/oregon

leff@smu.UUCP (Laurence Leff) (04/22/88)

TECHNICAL REPORTS FOR 1985

Computer and Information Science Department
University of Oregon
Eugene, OR  97403


CIS-TR-85-01 Automating the Transformational Development of Software
by Stephen Fickas.

Abstract:  This paper reports on efforts to extend the Transformational
Implementation (TI) model of Software Development [1].  In particular, we
describe a system that used AI techniques to automate major portions of a
transformational implementation.  The work has focused on the formalization
of the goals, strategies, selection rationale, and finally the transformations
used by expert human developers.  A system has been constructed that includes
representations for each of these problem solving components, as well as
machinery for handling human/system interaction and problem solving control.
We will present the system and illustrate automation issues through two
annotated examples.  This report will appear in IEEE Trans. Soft. Eng. 1985.

CIS-TR-85-02  AND Parallelism and Nondeterminism in Logic Programs
by John S. Conery and Dennis F. Kibler.

Abstract:  This paper defines an abstract interpreter for logic programs
based on a system of asynchronous, independent processors which communicate
only by passing messages.  Each logic program is automatically partitioned
and its pieces distributed to available processors.  This approach permits
two distinct forms of parallelism.  OR parallelism arises from evaluating
non-deterministic choices simultaneously.  AND parallelism arises when a
computation involves independent, but necessary, subcomputations.  Algorithms
like quicksort, which follow a divide and conquer approach, usually exhibit
this form of parallelism.  These two forms of parallelism are conjoinly 
achieved by the parallel interpreter.

CIS-TR-85-03 Polynomial time solutions of some problems in computational
algebra by Katalin Friedl and Lajos Ronyai.

Abstract:  The first structure theory in abstract algebra was that of finite
dimensional Lie algebras (Cartan-Killing), followed by the structure theory of
associative algebras (Wedderburn-Artin).  These theories determine, in a 
non-constructive way, the basic building blocks of the respective algebras
(the radical and the simple components of the factor by the radical).  In view
of the extensive computations done in such algebras, it seems important to
design efficient algorithms to find these building blocks.

We find polynomial time solutions to a substantial part of these problems.
We restrict our attention to algebras over finite fields and over algebraic
number fields.  We succeed in determining the radical (the ``bad part'' of the
algebra) in polynomial time, using (in the case of prime characteristic) some
new algebraic results developed in this paper.  For associative algebras we
are able to determine the simple components as well.  This latter result
generalizes factorization of polynomials over the given field.  
Correspondingly, our algorithm over finite fields is Las Vegas.

Some of the results generalize to fields given by oracles.

Some fundamental problems remain open.  An example:  decide whether or not a
given rational algebra is a noncommutative field.

CIS-TR-85-04 Building Control Strategies in a Rule-Based System by
Stephen Fickas, David Novick, and Rob Reesor.

Abstract:  Rule-based systems in which conflict resolution occurs require a
control strategy.  
This paper shows that effective conflict resolution often requires
domain-specific knowledge.  Moreover, defining control knowledge may be as
difficult as defining domain knowledge.  We present the ORBS model and
implementation of a rule-based systems environment which supports complex
and dynamic control strategies, and the interactive development of these
strategies.  ORBS and earlier systems are compared with respect to language,
control, and environmental support for development of control.  Four examples
of control using ORBS are discussed:  emulating a conventional system,
separating domain knowledge from scheduling knowledge, using an agenda-based
strategy, and dynamically changing a strategy during execution.  The ORBS
environment provides machine-aided construction of scheduling strategies
through a catalog of scheduling functions from which strategies can be
constructed, a skeleton scheduler, a scheduling editor, a scheduling tracer,
and a break package.  Finally, the paper discusses a model for system 
implementation of control from user-provided examples.

CIS-TR-85-05  Control Knowledge in Expert Systems:  Relaxing Restrictive
Assumptions by Stephen Fickas and David Novick.

Abstract:  Many expert systems use fixed, domain-independent control 
strategies.  These systems often embody assumptions about control and control
knowledge which limit the utility of application programs.  We present
the ORBS model and implementation of a rule-based systems environment which
attempts to expose, and give users access to, hidden control points. ORBS
provides a language for defining control strategies, and supports the 
interactive
development of these strategies.  We discuss the process of explicating control
knowledge through a process of refining an application rule.

CIS-TR-85-06 Design Issues in a Rule-Based System by Stephen Fickas.

Abstract:  This paper discusses the interaction between language, 
environment and
development model in the realm of rule-based systems. In particular,
we discuss our efforts in building a rule-based language and development
tools that support incremental development, interactive debugging, and
reuse of components.

CIS-TR-85-07 Computing the Composition Factors of a Permutation
Group in Polynomial Time by Eugene M. Luks.

Abstract:  Given generators for a group of permutations, it is shown 
that generators for
the subgroups in a composition series can be found in polynomial time.  The
procedure also yields permutation representions of the composition factors.

CIS-TR-85-08 The Concave Cusp as a Determiner of Figure-Ground by Kent A.
Stevens and Allen Brookes.

Abstract:  One method by which figure is distinguished from ground is on the
basis of information provided by concave cusps in contours. Concave cusps are
prevalent in natural textures, occurring in images wherever the silhouettes of
convex, opaque objects about or partially occlude one another.  The points of
contact between silhouettes present concave cusps, each indicating the local
assignment of figure versus ground across the contour.  There are known
tendancies to interpret as figure those regions that are lighter, or smaller or
more convex.  We show that the concave cusp is another, distinct, determiner of
apparent figure-ground that provides a stronger influence on figure-ground than
lightness or size.  While concave cusps are associated with overlapping convex
figures, the salience of the cusp appears to derive from the local geometry and
not from the adjacent contour convexity.

CIS-TR-85-09  Fast Parallel Computation with Permutation Groups
by Eugene M. Luks and Pierre McKenzie.

Abstract:  We develop fast parallel solutions to a number of basic problems
involving solvable and nilpotent permutation groups.  Testing solvability is
in NC, and RNC includes, for solvable groups, finding order, testing
membership, finding the derived series and finding a composition series.
Additionally, for nilpotent groups, one can, in RNC, find the center, a
central composition series, and point-wise stabilizers of sets.  There are
applications to graph isomorphism.  In fact, we exhibit a class of
vertex-colored graphs for which determining isomorphism is NC-equivalent to
computing ranks of matrices over small fields.  A useful tool is the 
observation
that the problem of finding the smallest subspace containing a given set of 
vectors and closed under a given set of linear transformations (all over a
small field) belongs to RNC.

CIS-TR-85-10 An Environment for Building Rule-Based Systems: An Overview
by Stephen Fickas, David Novick, and Rob Reesor.

Abstract: This paper presents an overview of the Oregon Rule-Based
System (ORBS), an environment which supports the construction and
development of domain knowledge and control knowledge for rule-based
expert systems.  We describe the operation of the system, with
particular attention to issues of control knowledge.  The paper then
presents an interactive model for development of application systems.
Finally, the paper discusses KATE, a knowledge-based, problem-solving
specification language and its use in constructing expert systems.

CIS-TR-85-11 Orthogonality in the Design of Command Languages by
Cathryn A. Stanford, Edward M. Bowden, David Locke, and Sarah Douglas.

Abstract:  Research has shown that organization plays an important role in
memory.  The present study applies these findings to the design of a command
language.  The concept of orthogonality is used to maximize the internal
organization of a text editing command language.  This orthogonal language is
compared to an organized, but non-orthogonal, and an anti-organized language
on measures of predictability, recall, and performance.  Subjects in the
orthogonal language condition performed better than subjects in the other
conditions on all measures.  General steps necessary to design an orthogonal
language are discussed.


CIS-TR-85-12 Zero divisors and invariant subspaces by Lajos Ronyai.

Abstract: In this paper we continue the study of algorithmic problems
related to associative algebras we initiated in recent work by
Friedl-Ronyai.  The main result of this paper is an algorithm to find
zero divisors in associative algebras over finite fields.  The idea of
this algorithm comes from an almost constructive proof of
Weddernburn's theorem on finite fields given in Herstein's algebra
text.

The above result is applied to the problem of finding common invariant
subspaces for a set of matrices over a finite field.  Using this
method one can give a polynomial time algorithm to find certain
minimal normal subgroups in permutation groups.  The corresponding
questions over fields of characteristic zero remain open.


CIS-TR-85-13 A Knowledge-Based Approach to Specification Acquisition
and Construction by Stephen Fickas.

This work is concerned with the automation of the software
specification process.  Specification automation research to date has
concentrated on non-interactive translation of complete, correct but
informal problem descriptions into formal specifications.  The work
proposed here is unique in that it addresses the construction of
complete, correct informal problem descriptions from incomplete, often
incorrect user descriptions. The proposed system will rely on
interactive problem solving and large amounts of domain specific
knowledge to carry out this process.  The final output of the system
will be a formal specification; our current target is the Gist
specification language, although other languages are under
consideration.