Word Processors
, DataBase Management Systems and Spreadsheet programs can all be viewed as information processing systems. Data can be organized with and transposed between the various programs, (although an entire data set is in reality rarely transposed). One can view data organization from three different points: abstract, logical and physical. The abstract view is the data organization perceived by the user of the program. The logical view is the data organization implemented internally in the application, (as created by the programmer). The physical view is the actual hardware data organization.Each of the following sections will begin with a brief description of the program, the typical functions it provides and the type of data organization underlying the program. The last section, Data Transformation, will depict how data can be reorganized into the underlying storage scheme of the various programs.
WYSIWYG, (What You See Is What You Get ), word processors, (eg. MS Word), are actually a combination of two related programs: a text editor and a text formatter.
Text Editors, (used extensively in programming), provide character and string, (a group of related characters), entry and manipulation operations such as insertion, modification, search and replace, etc. Text editors, (eg. UNIX EMACS), usually operate with fixed-size, (non-proportional or mono-spaced ), fonts in which each character is alloted the same horizontal space. This results in characters being aligned into rows and columns.
Text formatters, (eg. TeX / LaTeX, nroff
/ troff, etc.), accept an input file which contains not only the actual text of a
document, but also layout and typesetting of the document. This includes such things as
the titles, headings, footnotes, margins, justifications, paragraph delimiters, type
style, (bold, italics, strikethru etc.), page numbering,
etc. Text formatters typically use variable size, (proportional ), spaced fonts, in
which each character may be given a different amount of horizontal spacing. This results
in characters NOT being aligned into columns and equal sized rows. When output is
displayed on a screen the formatter attempts to show a result as a close as possible to
the final hard-copy output, (as WYSIWYG word processors).
The abstract storage organization for word processsors is based upon a written document. As such they have operations to support traditional documents, (eg. content tables, headers, footers, paragraph delimiters, outlining, indexing and a host of others). This allows, (and for the most part requires), users to organize their information into this paradigm when utilizing a word processor.
The underlying logical storage organization of word processors is that of a sequential list of symbols, (stream of characters, words or lines).
The actual computer memory (RAM) storage representation of this list will differ depending upon the particular word processor.
Database systems are best defined in terms of their position and relation to the standard business data grouping:
The majority of database systems use a relational organization, (eg. MS FoxPro). In this scheme data is stored as a set of hierarchically related tables. Users perceive the organization abstractly as tables, (i.e. files), which within the same database set must share a common field, (column), with at least one other table in the database, (as can be seen in the FoxPro example).
A discussion of the underlying logical storage organization of database systems is, as the saying goes, beyond the scope of this text. Suffice it to say that the underlying logical structure, (Btrees), are hierarchical allowing a quasi-direct mapping to the abstract relational table display. Just as the use of any tool, (informational or otherwise), has direct affects upon it's users, so do databases. Users must organize their data hierarchically in tables in order to be effective.
In addition to the obvious support for the hierarchical organizational paradigm, databases provide operations allowing users to re-organize data into answers to questions, (i.e. queries), reports, subsets of the the fields, designing custom input screens, etc.
Spreadsheets, (eg. MS Excel), are the electronic realization of an accounting ledger. They allow a user to abtractly organize data, (mostly numeric), into a column and row format and specify the mathematical relationships that exists between the various cells, (i.e. the intersection of a row and column - specified by the column letter followed by the row number or the cell label). Spreadsheets are often used to setup discrete simulations. By changing key base values that are related to many other values in the spreadsheet a user can quickly view the simulation affects due to the immediate recalculation of all cells.
A user of a spreadsheet logically organizes their data into a two-dimensional matrix of static, (non-computed), values and formulas for the dynamic, (computed), values. This organization forces a user to decide upon a two-dimensional prioritizing of their data. A person must be concerned with the layout of the data in the matrix due to it's potential affect upon rapid comprehension of the information. The following image shows a "Break-even Analysis" MS Excel spreadsheet displaying the formulas that have been entered.
The physical representation of spreadsheets would involve a discussion of the mapping of a matrix onto the one-dimensional, linear, memory of a computer. If interested the reader is referred to Pascal Plus Data Structures, third ed. by N. Dale & S. Lilly, pages 129-134 from D.C. Heath and Co. © 1991.
Despite the specific uses that the common computer applications have been designed around, one should realize that data is fluid. The organization of data into information can be altered by one to best match a particular problem at a given moment. Often new insights and understandings of complex data can be derived by transforming it into different underlying organizations.
Consider the following simple example of student's grades in a course, (the numbers are purely fictional). In a word processor the data might be represented as a simple list:
021542353 73 75 77 72 71 64; 027589848 87 97 100 88 91 87; 054648271 82 79 82 93 83 78; 141700439 91 50 93 89 81 79; 145725616 90 100 95 63 80 66; 164648428 79 67 38 50 48 27; 215824220 88 99 84 83 83 73; 220766654 87 76 68 70 54 46; 220780125 90 95 75 80 77 60; 224134136 92 92 75 80 81 73; 224339385 84 62 83 84 76 69
This organization clearly shows the order of each piece of data, (i.e. Soc Sec Number, Program 1, Program 2, Program 3, Test 1, Test 2, Test 3. However the organization explains little if anything about relationships in the data. In a tabular spreadsheet organization, the relationships in the data can more clearly be represented:
While this organization depicts more relationships, it does not clearly show the
derived nature of the data. Reorganizing into a hierarchical database, (tree),
organization allows one to quickly see this interdependency of the data:
