Our Office

Our company is a software house whose main business is developing computer program. Members are about one person, so to say it is a typical SOHO(Small Office Home Office), that is, my one-man company. We, too have gotten into financial difficulties since about 1993 on the same railway of the bubble collapse Japan. Nowadays we barely live from hand to mouth by my being temporarily transferred as a part-time lecturer of certain professional school.

Although it is four years since we began the project, yet our developing "Zelkova Tree" is not published. Zelkova Tree is a high-level software to be fit for specialists, but for the present we intend to merchandise it for general end users.

• Laboratory Address
• Company Resume
• Our Schedule
• Research Works
• Patent Retaining
• Japanese edition

Laboratory Address

Now, we rent a room in the northern countryside of Kanto plain, about 10 minutes on foot from Fukaya Station of JR Takasaki-line and are advancing our development and research work. Please visit us when you come nearby.

company: baba laboratory

address: 1-3-72-2H Inaricho Fukaya City, Saitama Pre., Japan 366-0026

Tel: (81)48-574-3585

home page: http://www.geocities.com/babalabo/

mirror site: http://www.aya.or.jp/~babalabo/

e-mail: babalabo@zelkova-tree.net

Company Resume

September, 1999

Our Schedule

Research Works

Our research works at present are congregated to the three clauses below.

• develop Directed Acyclic Graph Input-Output System (Zelkova Tree)
• design and implement Abstract Graph Experiment System (Monoverse)
• study Hamiltonian Circuit Problem (NP-complete Problem)

What came to start our learning is the development of mathematical expression input system as mentioned in the page of Patent Retaining. Mathematical expression (or structural document having grammar and notation) can be represented as tree in graph theory. That is, it can be said that taking tree-structure data as the simplest graph, we entered from the problem how can we make their input-output user-friendly, in other words, from the problem of user interface and visualizing.

We developed by necessity a multiway-tree data structure which we call N-tree. N-tree is a pointer array from the view of programing, and this conception is apparently seen under the influence of the idea of slots of M.L. Minsky. It is known that arbitrary graph has spanning trees. In this sense, N-tree is a spanning tree of the objective graph. The specification of N-tree is very primitive and it includes even how to locate data on physical memory.

Regarding N-tree, we can designate following characteristics.

(1) Every object is derived from Nodule class which is the base class of N-tree.
(2) Except root node, every nodule has double link with the parent node.
(3) With the double links, tree structure is generated spontaneously.
(4) General graph can be represented by dealing with non-tree-branch links called reference.
(5) N-tree is able to be serialized in general.
(6) It means that N-tree is able to be dealt as persistent object.
(7) Although Nodule is a kind of array class of C++, each slot is able to be named.

What we tackled next to the mathematical expression input system is simulation system principally for science education of middle/high school. This is the applied version of our N-tree technic, and it has such characteristics as (1)programming in mathematical expression, (2)visualization by animation, (3)description of parallel process with frames, (4)automatic calculation on constraint formula, (5)run time programming

It can be said that the genealogy software "Zelkova Tree", which is the application currently we are working, is leveled up with objective graph from tree to directed acyclic graph standing upon these practice. As it is able to concern that directed acyclic graph is a simple graph next to tree, it can be said that we have proceeded quite steadily by step.

Marriage is possible to be realized even if there is so much generation gap, however it is impossible that a node at the position of blood ancestor lineage comes to a child of the descendant lineage, therefore genealogy data makes a directed graph with no cycle, that is, DAG(directed acyclic graph). The problem to display genealogical tree or genealogical forest regards topological sorting mentioned by D.E. Knooth in Literate Programming. The algorithm itself, transforming semi-ordered set to liner order, might be short enough but it has another hardship in the view of concrete layout problem.

On the development of abstract graph experiment system, such two demands are concerned as (1)the demand which comes from the lower of these application development and (2)the demand which comes from the upper of experimentation of Hamiltonian Circuit Problem. Application development requests data verification tools for debugging in various stages, and theoretic study requires practical tools for sample generation or algorithm inspection. Indeed, it is almost impossible to chase a graph of more than 20 points with the naked eye. When we debug genealogy software, we have to get into the data structure nesting deeply and to trace pointers containing several stairs of indirect reference.

We have an conception to build up this abstract graph experiment system on brand new basis. Or, it might be better to paraphrase that to re-structure it as brand new basis. Although we describe every application using N-tree or Nodule class at present, we will describe all of our applications by the method which we call now Monoverse in the future.

To solve Hamiltonian Circuit Problem or NP-complete Problem is the subject assigned to us. We have a mind to undertake it. What we are mostly at a loss to make out the meaning is why such problem was turned over to us, not to others. However at least "pleasure" is given to us by the issue. We hope to enjoy it very much.

Patent Retaining

Our most basic technology is what we call N-tree. This technology was established in the process of developing Mathematical expression input system for CEC(Center for Educational Computation) ordered by Matsushita Communication Industry Co.. With regards to the technology, we retain the U.S. Patent as below.

• Patent Number: United States Patent No. 5,572,642
• Date of Patent: Nov. 5, 1996
• Title: DATA INPUT DISPLAY SYSTEM FOR PREPARING AND EDITING STRUCTURAL DATA
• Inventor: Michiro Nasu, Fukaya, Japan
• Assignee: Baba Laboratory Inc., Saitama ken. Japan
• Appl. No.: 237,108
• Filed: May 3, 1994
• ABSTRACT:

A data input display system makes it possible to input data which have a document structure having constraint in disposition of constituent elements. Such a structure is used in mathematical formulae, chemical formulae, and other data which have grammar and notations. The data input display system allows this type of data to be entered with a same level of ease and simplicity as the input of ordinary characters, and can ensure correct display and syntactic correctness of an input document; the system automatically configures the data on the display screen, based upon an integrated input, construction, and configuration process.

On the occasion of the application of this U.S. Patent, we received both physical and spiritual sustainment from the vice ex-president Hiroyuki Mizuno of Matsushita Electric Industrial Co. and the said Intellectual Property Center. We put it down here and express our gratitude from the bottom of our heart.