Speaking of characters and binary representation,
I was bored yesterday
(rainy vacation) and wrote
a bogus patent application on the ASCII table.
This
might be a forum where people appreciate it.
Or not. But here
goes.
Signature bogus patent traits of this include:
- Claim
something that is trivial, but make it sound difficult.
- Claim something
that is obvious to anyone with any skill in computer science.
- Claim
something that has been in common use for ages, and fail to mention relevant
prior art.
- Claim something that is smply a computerized version of a
very ordinary process.
- Laboriously use uncommon terms or invent new
words instead of calling anything by its common name.
- Claim incredibly
complex processes as if they were simple
extensions.
Method and apparatus for representing
numero-alphabetics in an analytical engine: encoding, storage, processing and
retrieval
Modern analytical engines commonly operate according to
the bi-valued principle and represent data as a bi-valued digits, henceforth
referred to as "bigits". One bigit can be interpreted as either the
complementary logical values "true" or "not true", or the numbers zero (0) and
one (1). To represent numbers other than 0 and 1, a multiplicity of bigits are
commonly employed in succession, as presented by several teachings in prior
art.
Numero-alphabetics is defined as a bounded but potentially large
ordered set of alphabetic letters, numerical symbols, diacritical marks and
other symbols, said ordered set of symbols being used for placing into storage
or presenting for processing or communication an arbitrary long but bounded
passage of language, the definition of language including at least human
language, computer language and language of hitherto undiscovered language-using
intelligent or semi-intelligent entities of biological or synthetic origin,
terrestrial or extra-terrestrial, as well as language of entities with a
communicative role but possessing only simulated intelligence.
Our
method and apparatus for representing numero-alphabetics utilizes the prior art
observation that a bounded group of N bigits can be employed to represent any
whole number, henceforth referred to as "non-fractional quantity", of value no
less than zero and no greater than the number two taken to the mathematical
power of N, less 1.
By our non-trivial, novel and inventive method,
these non-fractional quantities from a strictly bounded set of pre-negotiated
values can be uniquely mapped through an apparatus, the "correspondence vector"
(see Fig. 1) in a manner such that each non-fractional quantity corresponds to
no more than one numero-alphabetic symbol in the vector. A correspondence needs
not be established for every non-fractional quantity, i.e. the mapping needs not
be strictly injective nor surjective, provided that the non-fractional
quantities which do not correspond to a numero-alphabetic symbol are not used,
and that a mapping is defined at least for the subset of numero-alphabetic
symbols required for representing the particular text and the particular
language of the application at hand.
By deciding on such a
pre-negotiated correspondence vector between non-fractional quantities and
numero-alphabetic symbols, an ordered and bounded sequence of such
non-fractional quantities comprise a sufficient method for representing text as
bigits.
Section 2A: Encoding
In one embodiment, the
invention comprises a multiplicity of switches, possibly but not necessarily
arranged in the manner of a conventional typewriter for reasons of familiarity
to a human operator, where each switch is connected to one line of an encoder
circuit of logic gates, the closing of one switch creating a pattern of bigits
on a multiplicity of outputs, said multiplicity typically but not necessarily
representing a smaller multiplicity than the multiplicity of switches. The
correspondence vector is implicitly encoded in a combination of several discrete
but interdependent design details of the apparatus: i) the routing of individual
switches to the input lines to the encoder circuit of logic gates, ii) the
functional design of the circuit of logic gates, and iii) the enumeration of
each of the outputs of the circuit of logic gates. Care needs to be taken to
design all these non-trivial parts of the apparatus in conjunction to make the
embodiment of the invention function as intended.
Section 2B:
Retrieval
In one embodiment, the invention comprises a read-only
memory (ROM) whereby the input sequence of bigits is applied to the address
lines of said ROM, and a corresponding, typically but not necessarily longer,
output sequence of bigits is retrieved from the data output lines of said ROM,
the output forming a pattern of zeroes and ones that, when arranged in a
pre-defined ordered rectilinear grid, represent a discretized graphical image of
the glyph suitable for presentation on a digital indicator panel with a
rectilinear arrangement of individual visualization elements, each with a
human-interpretable visual representation of the two states "on" and "off". One
possible embodiment of such an indicator panel would comprise an array of
lights, where "on" and "off" would be visualized as "light" and "no light". Note
that we do not claim all embodiments of such indicators, we only claim their use
for visualization of numero-alphabetic symbols by means of a correspondence
vector based on a representation as non-fractional quantities encoded as bigits.
The correspondence vector is implicity encoded in the ROM as a correspondence
between a non-fractional quantity and a human-interpretable pattern of bigits.
The human-interpretable pattern of bigits is not primarily intended for
interpretation by an analytical engine, but such interpretation is possible. We
claim such an application as an extension of our original claim. This would
facilitate conversion to the native representation of the analytic engine when
that representation has been lost, or when automated interpretation of written
text is beneficial, regardless whether such text is typeset and printed, created
by means of a typewriter or potentially hand-written by a
human.
Section 2C: Storage
In one embodiment, the
invention comprises a methodology of using a fixed length sequence of bigits for
representing the non-fractional value corresponding to a certain symbol. In such
an embodiment, the sequence length can be, but need not be, chosen as an even
multiple or divisor of the otherwise natural design choice for grouping bigits
together to form larger sets of bigits in the analytical engine whereupon the
apparatus is based.
Section 2D: Processing
While
encoding, storage and retrieval of language by means of an analytical engine is
in itself useful and novel, the nature of analytical engines is that information
can be processed according to a pre-defined machine-interpretable set of rules,
an "operational sequence". While the construction and operation of such rules
are in the prior art, their applications to operating on written language are
enabled by the claims of this patent and tightly dependent on our claimed
inventive step, namely the correspondence vector. We therefore claim all
analytical engine operations on written language encoded by our claimed method
as contained in our teaching of using a correspondence vector. Possible
operations include, but are not limited to, counting the number of letters in a
written passage, comparing the groups of symbols against a pre-stored assembly
of allowed groups of letters to detect language mistakes, translation from one
language (human or otherwise) to another (human or otherwise), detection and
deletion of unwanted or disallowed words and phrases from a digitally encoded
passage of language, e.g. literature, and replacing all occurrences of a certain
symbol or sequence of symbols with another sequence, possibly of zero
length.
[ Reply to This | Parent | # ]
|