Visual communication device4503426Abstract A visual communication device for displaying combinations of written words which are entered by manipulation of an efficiently organized keyboard. The keyboard has certain keys specifically dedicated to a set of the words recurring most frequently in common conversational language as well as keys corresponding to discrete letters of the alphabet for spelling less-frequently used words. The keyboard is backed up by an extensive list of common terms stored in a memory. Each word is automatically fetched from the memory and displayed on a multi-digit alpha-numerical readout as soon as a number of leading characters sufficient to identify the word has been entered; thus, dispensing the operator from keying the remaining characters. In this first version, the device can be used by a person with auditory or vocal disability to quickly enter phrases in the readout by a much lesser number of manipulations than would be required if each word had to be spelled one letter at a time on the keyboard. Claims What is claimed is: Description FIELD OF THE INVENTION
__________________________________________________________________________
Predicate Phrase
__________________________________________________________________________
(a)
SHIFT the man
SHIFT DOES NOT LIKE TO
SHIFT work in the shop.
(b) YOU DO SHIFT swim well.
(c) SHE DID NOT WANT TO DO IT.
(d)
SHIFT a red house.
(e)
SHIFT the boy-s
SHIFT WERE NOT GOING TO HAVE IT.
(f)
SHIFT that dog
SHIFT WHEN IS IT SHIFT sleep-ing here?
(g) WHY DID THEY HAVE TO SHIFT rob the store?
(h)
SHIFT the ugly ghost
SHIFT SHIFT in the movies.
(i)
SHIFT this little girl
SHIFT SHIFT walking slowly.
SHIFT SHE COULD NOT BE
SHIFT go-ing to my house.
(j)
SHIFT my father
SHIFT HAS SHIFT read your article and
SHIFT SHIFT SHIFT taken notes.
(k)
SHIFT my father
SHIFT HAS SHIFT read your article in.
SHIFT a professional
newspaper
__________________________________________________________________________
The device data bases, i.e., the vocabulary lists, are stored in a series of EPROM's. This data base is subdivided into the two sections: the Main Vocabulary Storage (MVS), and the Keyboard Vocabulary Storage (KVS), which has a 256-byte/section dedicated to the characters, numbers, words and their combinations which can be directly entered by the actuation of a single key. In order to expedite the fetching of words from the data base and to minimize the storage requirement, the alphabet is divided into two groups of letters. Those letters which are most frequently used (A, C, D, H, I, L, M, N, O, R, S, T, N and U); and those which do not recur as often in the English and Spanish languages (B, F, G, J, K, P, Q, V, W, X and Y). Only four bits are used to represent the letters in the first group. One letter from each group can thus be stored in a single 8-bit address location. The hexadecimal code F is used to indicate that the following half-byte contains a letter of the second group. The device is programmed to interpret each half byte as a letter of the first group unless it is preceded by hexadecimal F. A zero is used to indicate the end of a word. It is estimated that this compact storage method yields a thirty percent savings on memory space. Table A illustrates the addressing and coding arrangement of a section of the KVS.
TABLE A
______________________________________
KEYBOARD VOCABULARY STORAGE (KVS)
BYTES CORRESPONDING
WORD Address Contents LETTERS
______________________________________
ESTA 3C 0 3 E
3D A B S T
3E D 0 A
ARE 3F D 9 A R
40 3 0 E
ES 41 3 A E S
IS 42 0 5 l
43 A 0 S
LA 44 6 D L A
THE 45 0 B T
46 4 3 H E
A 47 0 D A
TO 48 0 B T
49 8 0 O
DE 4A 2 3 D E
OF 4B 0 8 0
4C F 3 F
ESTABA 4D 0 3 E
4C A B S T
4F D F A
50 1 D B A
WERE 51 0 F
52 7 3 W E
53 9 3 R E
UDS AD C 2 U D
AE A 0 S
YO B2 F 9 Y
B3 8 0 O
I B4 5 0 I
USTED B5 C A U S
B6 B 3 T E
B7 2 0 D
YOU B8 F 9 Y
B9 8 C O U
ELLA BA 0 3 E
BB 6 6 L L
BC D 0 A
SHE BD A 4 S H
BE 3 0 E
ELLOS BF 3 6 E L
C0 6 8 L O
C1 A 0 S
THEY C2 B 4 T H
C3 3 F E
C4 9 0 Y
______________________________________
It should be noted that words are grouped in the KVS memory according to their functions. Thus, the address of the word itself indicates whether it is a verb or a pronoun. It should also be noted that words are stored adjacent to their translation. The addresses of the words stored in the KVS are grouped in a separate section of the data base called Keyboard Lookup Table (KLT). In the KLT, addresses of all the characters, letters or words which may be entered from a single key and their possible translation are grouped together. Table B illustrates the organization of a section of the KLT. The addresses of the letter P and that of the various forms of the verb "to be" (and Spanish translation), which can be entered by means of a single key, are held at the locations 47 through 50. The processor 5 has the ability to select among these entries the one which must be displayed according to the circumstances.
TABLE B
______________________________________
KEYBOARD LOOK-UP TABLE (KLT)
Corresponding
Address Contents Word
______________________________________
3 D C D
3 E F 4 P
3 F 5 1 WERE
4 0 6 A WAS
4 P 6 A WAS
4 2 6 8 ERA
4 3 6 8 ERA
4 4 4 D ESTABA
4 5 4 D ESTABA
4 6 0 1 SER
4 7 E D
4 8 0 5 I
4 9 3 F ARE
4 A 4 2 IS
4 B 6 6 AM
4 C 4 1 ES
4 D 6 4 SOY
4 E 3 C ESTA
4 F 1 1 ESTOY
5 0 0 1 BE
______________________________________
Fetching from the MVS area of the data base of words which are spelled, character-by-character, using the lower case entry mode, follows a different procedure. After spelling one or more leading characters of the desired word, the operator presses the STEP key. At this point the keyed data, held in a special register called Letter Processing Area (LPA) is sequentially compared to the words stored into the MVS. As soon as a coincidence is detected between the LPA contents and the leading character of a MVS word, that word is fetched and displayed. If the displayed word does not correspond to the desired one, the operator presses the SPACE key, causing the device to step through the remaining part of the MVS until the next matching work is found. The process is repeated until the readout displays the desired word. The operator can then proceed to the next key entry without touching the SPACE key. If the program reaches the end of the MVS without finding the desired word, an appropriate error message is displayed. The formating of the phrases and sentences and the selection of the proper conjugation forms which are necessary to complete a grammatically correct translation is done in an area of the memory 6 called the Sentence Pattern Storage (SPS). The SPS does not hold, or operate upon, the words themselves, but only their addresses. Accordingly, the SPS receives the word address fetched from the KLT and those obtained through the search and compare process described earlier in connection with the spelled words found in the MVS. The expanded block diagram of FIG. 4 illustrates the important role played by the SPS. As the operator enters a message through the keyboard 3, the KLT 10 is first addressed; and the address of the corresponding word is stored in the SPS 14. The SPS in turn, addresses the KVS 9 or MVS 8. The letters and words are then extracted from the memory and stored in the End of Word Storage (EWS) 12 which is associated with the alphanumeric readout 4. During the translation process, the SPS is loaded with the addresses of the target language words and causes these words to be extracted from the KVS, the MVS, or both, and to be stored in the Translated Sentence Storage (TSS) 13. The contents of the TSS appears on the readout 4 in place of the base language version. The KLT, KVS and MVS are all part of the Erasable Programmable Memory (EPROM) 6. The SPS, EWS, and TSS are parts of a Random Access Memory (RAM) 11. The operation of the SPS can be best understood by referring to Table C which illustrates its organization. The SPS is subdivided into twenty-one slots labeled A through U. Each slot is made from two bytes which suffice to hold the address of a vocabulary word and various indexes and codes which are used in the processing of that word. The three first slots are dedicated to the subject phrase; the first slot being dedicated to the pronoun or article, the second slot to the adjective, and the third to the noun. When the SPS contains the Spanish translation, the function of the second and third slots are inverted in order to accommodate the Spanish habit of putting the adjective after the qualified noun. Slots D through K are reversed for the predicate phrase. This section of the SPS holds addresses of words which are entered via the word keys. It should be noted that the English version can hold up to three auxiliary verbs. This provides sufficient flexibility to construct such phrases as "she did not want to do it38 or "they were not going to have it". The Spanish version can be written with the use of a lesser number of auxiliary verbs. Accordingly, the slot J is used to store the address of the preposition "a", or the addresses of the suffixes for the third and second person plural forms of verbs, as might be required. Slots L through T are reserved for storing the addresses of the words forming the object phrase. Slots T and S are reserved for words which are either found in the MVS or are entered letter by letter under control of the DIRECT key and which are not translated; but are transposed directly into the translated sentence. These slots also receive the overflow from the object phrase. Slot U is used for the punctuation and holds the address of either the period or question mark.
TABLE C
__________________________________________________________________________
SENTENCE PATTERN STORAGE (SPS)
ORGANIZATION
RAM English
Spanish
Slot Address
Function
Function Examples
__________________________________________________________________________
Subject
A 30
31 Pronoun or
Pronoun or
THE LOS
Phrase Article
Article
B 2E
2F Adjective
Noun BIG CABALLOS
C 2C
2D Noun Adjective
HORSE-S
NEGROS
Predicate
D 2A
2B Pronoun
Pronoun THEY
Phrase
E 28
29 Adverb Adverb
F 26
27 Auxiliary
Negative
ARE NO COULD
NO
G 24
25 it Lo
Pronoun
H 22
23 Negative
Auxiliary
NOT VAN NOT PODIAN
I 20
21 To Auxiliary
GOING BE ESTAR
Auxiliary
J 1E
1F To A -N TO A
Auxiliary
MOS
K 1C
1D Auxiliary
Preposition
Preposition
Article
or Article
Object
L 1A
1B In Or Verb RUN CORRER FLY VOLA
Phrase Verb
M 18
19
ING
NDO
ING
NDO
N 16
17 Preposition
Preposition TO A
O 14
15 Pronoun
Pronoun OUR NUESTRA
or Article
or Article
P 12
13 So-Very
Noun SO
Q 10
11 Adjective
MUY QUICK TAN CASA
or Adverb
TAN -s
R 0F
0F
LY or Noun
Adjective
LY RAPID
HOUSE
or Adverb
Extra S 0C
0D Adverb/
AMENTE AMENTE
IN EN
Words Conjunct
Adverb/Con-
Direct Word
junct -S or
S Direct Word
T 0A
0B Direct Word
Direct Word
NOW YA AUSTIN
AUSTIN
Punctuation
U 08
09 Punctuation
Punctuation
0 0 0 0
__________________________________________________________________________
The morphological and syntactic manipulations which take place in the SPS can be best explained in reference to the entry and translation of specific sentences. Assuming the English to Spanish mode of operation, let us consider the processing of the following sentence: THE BIG HORSES/ARE NOT GOING TO/RUN SO QUICKLY NOW--wherein "/" is used to separate the predicate phrase from the subject and object phrases. The three first words are immediately flagged as article, adjective and noun by reason of the place where they are found in the MVS. This, in turn, determines there SPS slot assignment, i.e., ABC, respectively. As the "T" key is pressed to enter the term "ARE" the procesor is directed to address location 47 in the KLT, then looks at the subject phrase to determine which of the two possible forms "AM" or "ARE" must be used. The absence of the pronoun "I" ahead of the auxiliary indicates that the KLT address should be incremented to 49 in order to select the form ARE. The terms "NOT", "GOING", and "TO" are the only English forms which can be entered in the upper case mode of their entry keys. Their respective addresses can thus be fetched directly from the KLT and placed in slots H, I and J. The term RUN could be either a noun or a verb. The processor examines the term which follows the past auxiliary verb in the predicate phrase to resolve the ambiguity. A lack of word or an article would indicate that the word "RUN" is a noun. The presence of the preposition "TO" however, always indicates that the following term is a verb to be entered in slot L. The adverb "SO" has a specially dedicated slot P. The following word can only be an adjective or adverb whose address is automatically stored in slot "Q". The suffix "-LY" is assigned to the next slot "R". The extra word "NOW" is recognized as an adverb and placed into the extra slot "S". The letters, symbols and words are fetched from the MVS memory and are loaded for display into the EWS as soon as their addresses are placed in their proper SPS slot location. The entry of the period in the punctuation slot U triggers the translation process. The translation of English words and combinations thereof is predicated upon the slot occupied by the English word addresses, the particular group of arrays where the words are stored, and any particular indicia which may be held within the SPS in order to indicate a special processing requirement. In the MVS section, the array which groups the nouns is subdivided according to the genders. Gender of the Spanish version of a word is always controlling, regardless of the fact that English version of an object or concept is neutral. Accordingly, "HOUSE:CASA" is stored with "GIRL:NINA" in the feminine sub-array; while "MAP:MAPA" is stored along with "BOY:NINO" in the masculine sub-array. The storage address of each noun thus automatically provides an indication of the Spanish version gender. In our example, the English article "THE" translates into Spanish as "EL" "LA" "LOS" or "LAS", depending upon the gender of the noun it qualifies. The address of the words "HORSES CABALLOS" indicates a masuline gender. The entry of "-S" after "HORSE" indicates the plural. Accordingly, the address of the form "LOS" is entered in slot "A". The plural forms of most nouns and adjectives is automatically formed in Spanish by tagging "-S" to the end-vowel of their singular form of "-ES" to the end-consonants of their singular forms. The feminine version of most adjectives is formed by substituting an "A" for the ending "O" of the masculine form. Nouns and adjectives with irregular plurals in English or Spanish are stored in a separate section of the MVS adjacent to their irregular forms. The same technique is used for adjectives with irregular feminine forms. The address of these words let the processor know that the irregular plural or feminine forms must be obtained by incrementing their address and counting the word separators until the proper form is reached. Table D illustrates the storage of nouns, adjectives and verbs with irregular forms. In the translation of the word "BIG" the address entered in slot C of the SPS for the Spanish version is that of the word BIG incremented until the third location corresponding to the proper irregular plural forms is identified.
TABLE D
______________________________________
Main Vocabulary Storage (MVS)
BYTES CORRESPONDING
WORD Address Contents LETTERS
______________________________________
GO B 0 1 0 F
2 0 8 G O
IR 3 0 5 I
4 9 0 R
GOING 5 F 2 G
(GERUND) 6 8 5 0 I
7 7 F N
8 2 G
YENDO B 0 9 0 F
(PAST B 0 A 9 3 Y E
PARTICIPLE)
B 0 A 7 2 N D
B 0 C 8 0 O
GONE B 0 D F 2 G
(PAST B 0 E 8 7 O N
PARTICIPLE)
B 0 F 3 0 E
IDO B 1 0 5 2 I D
(PAST
PARTICIPLE) 8 O
BIG 0 F
1 5
F 2 G
GRANDE C 0 1 0 F
C 0 2 2 9 G R
C 0 3 D 7 A N
C 0 4 2 3 D E
GRANDE C 0 5 0 F
(FEMININE)
C 0 6 2 9 G R
7 D 7 A N
8 2 3 D E
GRANDE 9 0 F
(PLURAL) A 2 9 G R
B D 7 A N
C 2 3 D E
______________________________________
Since the next term "ARE" was followed by "NOT" the address of the Spanish negative "NO" is entered into slot "F". The progressive form "TO BE GOING TO" is translated with the present and perfect tenses of the verb "IR". The phrase, "TO BE USED TO" as well as the auxiliary "TO DO" (as in I DO DRINK WINE) is translated with the present and perfect tenses of the verb SOLER. The KVS holds the Spanish present and perfect forms necessary to conjugate the following verbs: SER, ESTAR, HABER, IR, SOLER, HACER, TENER, and PODER. Those are the only conjugation forms available in the translation but which suffice to translate conversational English into Spanish within the restrictions previously outlined. In the example, the phrase "ARE GOING TO" thus, is translated with the present tense of "IR" and in taking into account the plural subject "CABALLOS", to yield the third person plural form "VAN". In order to economize memory space, only the form "VA" is stored in the KVS. The suffix "N" is added when the word is fetched from the memory for display. The SPS only holds the address of the form "VA" plus a code indicating the third person plural form. The infinitive "RUN" is translated into "CORRER" without special processing within slot "L". The three terms "SO, QUICK, and -LY" are also translated directly with "TAN" "RAPID" and "-AMENTE", but with a forward shifting of one slot position. The adverb "NOW" translates directly into "YA". Now, let us consider the entering and translating of the sentence "THEY COULD NOT BE/FLY/ING TO OUR HOUSE IN AUSTIN" into the unit. There is no subject phase, but only a personal pronoun whose address is held in slot "D". This address indicates that it is a third person plural form which translates into "ELLOS". The form "COULD" translates into "PODIA", plus suffix "-N" which is dictated by the third person plural of the pronoun. This is an adequate substitute for the form "PODRIAN". The negative is entered and translated directly, but placed after the auxiliary. The actuation of the "H" key to enter the infinitive "BE" presents to the unit a choice between the forms "IS" or "BE". The presence of the auxiliary "COULD" dictates that the latter is the proper form to enter. In the translation, the choice between "SER" and "ESTAR" is governed by the following gerund "FLY-ING" which requires the selection of "ESTAR". In the object phrase, the term "FLYING" must be entered in the lower case mode as "FLY" and the suffix "-ING". The address of this verb indicates that it is a regular verb whose past participle and gerund can be formed by substituting "-NDO" for the infinitive "R" form. The processor would fetch an irregular gerund such as "YENDO" from the MVS. The gender of the possessive in the translation of "OUR" is indicated by the address of the noun "HOUSE: CASA". The final phrase "IN AUSTIN" falls into the extra-word slots S and T. The preposition can be found directly in the memory; but the word AUSTIN must be entered manually using the DIRECT key. This mode allows the operator to enter non-translated word of up to ten characters. This feature is particularly useful in entering proper names. At the end of the translation process, the addresses stored in the SPS are used to fetch the words from the memory and load them into the translated sentence storage unit (TSS) 13 from which they are displayed on the readout 4. The techniques just described are used in the inverse process of entry and translation from Spanish into English. The schematic of FIG. 5 illustrates the interconnections between the keyboard switches 15, the processor modules 16 and 17, the EPROM modules 18, 19, 20 and 21, and the display control units 22, 23, 24 and 25. Module 16 contains the RAM memory 11. Module 18 holds the KLT and KVS. The MVS utilizes module 19, 20 and 21. The unit is powered from a series of batteries 26 or from a power supply 27 operating on household current. The program which is followed by the processor 5 is organized according to the flow diagram illustrated in FIGS. 6A through 6F. The program comprises seven major subroutines and 14 minor ones. Major subroutine A manages the entry of letters, words or control commands through the keyboard 3. Major subroutine B controls the processing of a sentence in the SPS after a punctuation mark has been entered. Major subroutine C handles the entry of the predicate phrase into the SPS. Major subroutine D controls the letter by letter entry of words which are not found on the keyboard. Major subroutine E processes the words stored in the EWS and then directs the program to accept a new input from the keyboard. Major subroutine F applies to the entry of letters through the keyboards. Major subroutine G controls the loading of the TSS according to the addresses found in the SPS. A four-position SHIFT COUNTER S is used to indicate the part of the sentence being processed. S=1 corresponds to the processing of the subject phrase; S=2 corresponds to the processing of the predicate phrase; S=3 corresponds to the processing of the object phrase; and S=4 indicates an error in the use of SHIFT commands. The term POS corresponds to the part-of-speech value given to a word as determined by the location of that word in the memory. The POS is derived by analyzing the address of the word in the KVS or the MVS. Flag X is set whenever a noun or pronoun is entered as the subject of a sentence. Flag Y is set when the first auxiliary verb is entered. This first auxiliary verb is always the one to be conjugated according to the person and number of the subject. Flag Z is reset while a yellow key or a suffix key is being processed. Flag W is set during a Spanish to English translation. Buffer A is used to hold the address of a spelled word fetched from the MVS or that of a yellow keyword or that of a word being entered in the predicate sentence. Buffer B works in conjunction with the LETTER PROCESSING AREA (LPA) to store the root of the word being spelled from the keyboard. Buffer C remembers the address of the last matching word found in the vocabulary and being displayed for acceptance or rejection by the operator. Buffer D stores the status of the subject which controls the conjugation of the auxiliary in the predicate sentence according to the following code. first person singular=11; second person singular=21; third person singular=31; first person plural=12; second person plural=22; and third person plural=32. The minor subroutines labeled SUB 2 through SUB 14 are not illustrated in the flow diagram but are described in Table E. These subroutines can be easily programmed into the device by a person skilled in the art of computer programming. The improper use of the unit results in the display of a variety of error codes labled E 03, E 04, E 10 through E 13, E 15 and E 16 whose meanings are given in Table F. The implementation of these error codes is shown in the flow chart of FIG. 6B (e.g. E 16 corresponds to too many shift commands). The addition of a communcation interface connected to the inputs and outputs of the ESS and TSS could be added to allow remote display of the entered and translated sentences and two-way conversations between two persons of different linguistic backgrounds, either by telephone or radio.
TABLE E
______________________________________
Minor Subroutines
______________________________________
SUB 2 STROBE DISPLAY
SUB 3 FIND KEY JUST INPUTTED
SUB 4 ERROR CODE DISPLAY
SUB 5 SCAN AND PULSE DISPLAY
SUB 6 (POS) DECIPHER WORD ADDRESS
SUB 7 (SAS) SLOT ASSIGNMENT PER:
POS VALUE, SLOT OCCUPANCY, AND
POSITION IN SENTENCE
SUB 8 CONVERT MEMORY LETTER CODE (4 BITS)
TO LED DISPLAY (8 BITS)
SUB 9 MOVE TRANSLATED WORD ADDRESS TO
APPROPRIATE LOCATION
SUB 10 WORD ADDITION OR DELETION
SUB 11 CONJUGATE ARTICLE OR ADJECTIVE
PER NOUN POS
SUB 12 PROCESS EXCEPTION TO RULE
SUB 13 TEST FOR VERB MODIFIER
SUB 14 ERROR SEARCH
______________________________________
While I have described the preferred embodiment of the invention and suggested modifications and additions thereto, it should be understood that other embodiments and applications to different languages could be implemented without departing from the spirit of this invention and the scope of the appended claims.
|
Same subclass Same class Consider this |
||||||||||