A patent (U.S. Patent No. 5,680,557, "Natural Computing System and Environment") describes an invention that allows people to do computations by using time-tested and trusted ways of representing data and information. This method relies on equations, tables, graphs, worksheets, unit strips, text, and so forth in computers as if these objects were on paper. In the "Natural Computing" environment, these natural features (equations, tables, etc.) are preprogrammed in object-oriented programming languages. Electronic documents based on the "Natural Computing" format allow people to use text and concurrently carry out computations. For the end user or the developer of a solution to a computational problem, no further programming is required, other than the manipulation of features in an obvious and instinctive way.

An example of a "natural Computing" screen.

The figure shows an example of a "Natural Computing"screen depicting a procedure object containing three other object types, namely a table, a graph, and a number of equations. Once the procedure object is highlighted and input values are provided by the user, result values are output. The procedure object and other objects are embedded along with the text in the electronic document.

The invention attacks the common software development problem in computation-intensive domains. Despite the enormous achievements in computation capabilities seen in the nineties, software development in calculation-intensive applications is still slow, costly, and unreliable.

This situation arises in large part because traditional software development methods require that subject-matter specialists explain their computational procedures to programmers before the data, knowledge, and procedures can be encoded for computer systems. The inherent complexity of current computer language schemes has led to the development of an army of professional programmers, analogous to the scribes who transcribed hieroglyphic symbols for use by ancient end users before the invention of widely accessible alphabets and Arabic numerals. The need for these intermediaries between the user and the task introduces inefficiency, incessant programming, and considerable lag between the availability of knowledge and its use.

The invention relies on the natural way people do computations. They read information from journals, textbooks, handbooks, encyclopedic references, brochures, and catalogs, then transfer that data, information, and methods onto paper; do the computation; and record the information resulting from the new computations. This invention consists of building classes that represent computational features such as equations, tables, graphs, worksheets, unit strips, text, and so forth in computers as if these objects were on paper.

The classes contain operations and methods to manipulate data in the objects that are instances of these classes. In a typical domain application, equations, tables, and so forth are represented as objects, and domain data and information are entered. For example, procedure objects are developed by connecting the necessary objects in the required sequence. The "Natural Computing" environment is intended to present a menu of objects for domain specialists to choose objects from as necessary for a given application.

The system presents numerous benefits. Authors of electronic textbooks, papers, and handbooks will be able to embed computational features and procedures in text for easy computation. Since users of these electronic documents will be able to use calculation procedures directly from these sources, the need for applications programming professionals will be eliminated. Avoiding these intermediaries increases efficiency and economy and decreases the gap between the generation of knowledge and its availability in electronic books. The impact of the "Natural Computing" system could be comparable to that of word processing, desktop publishing, CAD, spreadsheets, and database systems development.

This work was done by Som Karamchetty of the U.S. Army Research Laboratory. Inquiries concerning rights for the commercial use of this invention should be addressed to Ms. Norma Cammaratta, Office of Research and Technology Applications, Army Research Laboratory, AMSRL-CS-TT, 2800 Powder Mill Rd., Adelphi, MD 20783-1197; (301) 394-2952; fax: (301) 394-5818.

Electronics Tech Briefs Magazine

This article first appeared in the February, 1999 issue of Electronics Tech Briefs Magazine.

Read more articles from the archives here.