LaParser Module of AI4U Artificial Intelligence

Skip to first unread message

A.T. Murray

Jul 29, 2023, 6:13:30 PMJul 29
LaParser Latin Parsing Module of the
Mens Latina Artificial Intelligence in Latin Language

1. Diagram of Chapter 10 LaParser AI Mind-Module

/^^^^^^^^^\ LaParser Assigns Associative Tags /^^^^^^^^^\
/ EYE \ MINDCORE _____ / EAR \
/ \ CONCEPTS /New- \ / audMemory \
| _______ | | | | _____ (Concept)-|-------------\ |
| /old \ | | | | /Old- \ \_____/ | Audition | |
| / image \---|-----+ | (Concept)------|---|-----------\ | |
| \ recog / | | | | \_____/-------|---|---------\ | | |
| \_______/ | a| | | |________V | | | | |
| | b|C| | / LaParser \ | | | | |
| visual | s|O|f| \__________/ | | | | |
| | t|N|i| |noun? | DAT-----/ | | |
| memory | r|C|b| |verb? | | | |
| | a|E|e| |adj.? | DEUS------/ | |
| channel | c|P|r| |adverb? | | |
| | t|T|s| |prep.? | TIBI | |
| _______ | | | | |conj.? | | |
| /new \ | |_|_| ______V____ | PECUNIAM----/ |
| / percept \ | / Psy \ / \ | |
| \ engram /---|--\array/-----( InStantiate ) | |
| \_______/ | \___/ \___________/ | |

2. Purpose of the LaParser module

LaParser serves the purpose of not only identifying a part of speech such as a noun, preposition or verb, but also of comprehending the part of speech in context by helping to assign associative tags among concepts in the Psy conceptual array. Thus LaParser and its Russian counterpart RuParser and its English counterpart EnParser serve the purpose of Natural Language Understanding (NLU).

3. Algorithm of the LaParser AI Mind-Module

3.A. PraeScium bootstrap sequence relieves burden of parsing some words.

Gradually, as all the Latin conjunctions, prepositions and pronouns are encoded within the bootstrap sequence as separate parts of speech, the Latin parsing module simply recognizes them and does not need to parse them. However, recognition of a Latin preposition will help the parsing module to treat the associated noun or pronoun as the object of the recognized preposition.

3.B. OldConcept module prepares parameters for the LaParser() module.

Latin pronouns, being brief in length -- "ars longa, pronomen breve" -- may eventually be handled in their entirety by the OldConcept module.

3.C. Assigning associative tags to concepts based on inflectional endings

3.C.1. Initial proof of concept

The current LaParser() module first sends a concept into the InStantiate module and then almost immediately inserts associative tags into the flag-panel of each pre-instantiated concept. The process happens so quickly that for all intents and purposes it is like the swiftly operating, massive parallelism of the human brain. The immediate revisiting of the instantiation process is necessary only because the Latin AI Mind interprets a sentence of Latin input without regard to the word-order or syntax of the input, relying instead upon the inflectional endings of each Latin word. Since subject, verb and object (SVO) may be in any haphazard word order, the Latin parsing module must wait almost instantly for all the SVO words to be available for parsing into their grammatical roles.

The earliest forms of the LaParser module require input of unique nominative and accusative forms to function properly. As the Latin RoboMind matures, provisions must be made in the LaParser() module to disambiguate between Latin words which take the same form for more than one case, such as nominative and accusative case, as explained in the Latin AI User Manual.

3.C.2. Improving LaParser() to deal with ambiguous noun inflections.

[From the Mens Latina Programming Journal]
In the Mens Latina artificial intelligence in Latin language, we have made the LaParser Latin parsing module able to understand a subject-verb-object (SVO) sentence if the subject-noun and the direct-object-noun are unambiguously nominative or accusative. Now we want LaParser to deal properly with the input of Latin sentences in which one of the SVO nouns could be either nominative or accusative, and the parsing module must decide which is the case. We will use two similar sentences.

FRATRES HABENT PUERI. ("The boys have brothers.")
FRATRES HABENT PUEROS. ("The brothers have boys.")

The noun "fratres" can be nominative or accusative, depending upon whether it is the subject or the object of the verb. Since the verb needs a nominative subject, "pueri" is the subject in one sentence, and "fratres" is the subject in the other sentence. Our task in coding is to make the AI able to treat the right noun as the subject.

We have now set up some variables kbdba and kbnum to make the flag-panel values of a recognized concept available to the LaParser() Latin parsing module. These values are not the final determination of the case and number of a noun, because there may be some ambiguity. But they may serve as the default values when a recognized concept is first instantiated in Stage One of the Latin parsing.

In Stage Two, the LaParser() Latin parser module performs a series of tests to determine or disambiguate the speaker-intended case of the noun, based not only on its perhaps ambiguous inflectional ending but also on the other Latin nouns being used in the sentence of input. The rota counter allows the InStantiate() module to cycle or rotate through the several nouns that are possibly included in the Latin input, so that the OutBuffer() detection of potentially ambiguous Latin noun-endings may load the initial instantiation time-point onto unknown "upshot" placeholder variables to be kept ready during a series of "Stage Two" LaParser() tests in advance of re-instantiating each of the several nouns as either the tsj subject, or the tio indirect object, or the tdo direct object of the Latin verb. If the upshot variable (ux1-ux5; uy1-uy5; uz1-uz5) wins selection, it already contains the Stage One time-point of instantiation which will now be transferred to the time-point of the subject or indirect object or direct object, so that each item may be re-instantiated with the properly tested and identified associative tags which play their essential role in the Natural Language Understanding of the input sentence.

4. Code of LaParser() from Abra014A Mens Latina in JavaScript for MSIE

function LaParser() { //
// if (pos == 5) alert("LaP: psi= "+psi+" dba= "+dba+" num= "+num); // 2019-05-09: TEST
act = 48; // 2018-01-11: an arbitrary activation for InStantiate()
bias = 5; // 2018-01-10: Expect a noun until overruled.
if (pos == 5) bias = 8; // 2018-01-11: after noun, expect verb.
if (pos == 7) bias = 8; // 2018-01-11: after pronoun, expect verb.
if (pos == 8) bias = 5; // 2018-01-11: after verb, expect noun

Psy[toc].psyExam(); // 2010-05-10: examine flag-panel at time-of-old-concept
kbdba = psi7; // 2019-05-10: transfer dba-value to "kbdba"
kbnum = psi8; // 2019-05-10: transfer num-value to "kbnum"
kbmfn = psi8; // 2019-05-10: transfer mfn-value to "kbmfn"
dba = kbdba; // 2019-05-10: default value for initial instantiation.

InStantiate(); // 2019-05-05: for creating psy concept-nodes

// 2019-05-10: Stage Two
if (uy1>0 && uy4==0) { // 2019-05-10: if nominative but not accusative...
tsj = uy1; // 2019-05-10: if nominative only, then subject
Psy[tsj] = new // 2019-05-10: re-instantiate the noun as subject
} // 2019-05-10: end of test for noun nominative but NOT accusative

if (ux1 > 0 && ux4 > 0 && uy4==0) { // 2019-05-10: if form is both nom. and acc.
tdo = ux4; // 2019-05-10: accusative
Psy[tdo].psyExam(); // 2010-05-04: expose flag-panel of Latin direct object
Psy[tdo] = new // 2019-05-10: re-instantiate the noun as direct object
} // 2019-05-10: end of test for noun both nominative and accusative

if (tsj > 0 && tvb > 0 && tdo > 0) { // 2019-05-04: if SVO input complete...
Psy[tsj].psyExam(); // 2010-05-04: expose flag-panel of subject
Psy[tsj] = new // 2019-05-04: insert "tvb" as the "tkb" of the subject
Psy[tvb].psyExam(); // 2010-05-04: expose flag-panel of the verb
Psy[tvb] = new // 2019-05-04: insert "tdo" as the "tkb" of the verb
Psy[tdo].psyExam(); // 2010-05-04: expose flag-panel of the direct object
Psy[tdo] = new // 2019-05-04:
} // 2019-05-04: end of test for completion of subject-verb-object input.

} // 2019-04-26: LaParser() returns to OldConcept() or NewConcept().

5. Variables for the LaParser AI Mind-Module

act -- quasi-neuronal activation-level

bias -- LaParser(); NewConcept(): expected part-of-speech POS

dba -- doing-business-as noun-case or verb-person. For Latin verbs, LaParser() may use the b15 and b16 characters ending a verb in the OutBuffer() module to set the dba and prsn values for a verb about to be instantiated.

kbdba -- knowledge-base doing-business-as -- in the Mens Latina Latin AI, reveals what case a Latin noun is in at the particular toc (time of old concept) when the Latin noun was recognized as representing a particular concept. Because the inflectional ending of a recognized Latin word may be ambiguous with respect to, say, nominative case or accusative case, the value of the dba tag at the time of recognition is not the final say or final interpretation of what case a noun is in as part of a new Latin input. The value obtained from the toc node may be treated as a temporary default value likely to be overridden by the grammatical tests conducted by the LaParser() module.

kbmfn -- knowledge-base masculine-feminine-neuter -- in the Mens Latina Latin AI, is the gender (masculine/feminine/neuter) of a Latin noun as revealed at the toc time of its recognition as a word and a concept. The gender-value obtained for the noun is not likely to be overruled. Although in English a "friend" can be a he or a she, in Latin the gender is more closely bound with each particular noun.

kbnum -- knowledge-base num(ber) -- in the Mens Latina Latin AI, is the grammatical number (singular or plural) of a particular conceptual engram of a Latin word at the toc time of its recognition as a word and a concept. The given value serves as a default for the first instantiation of the input noun, and is not likely to be overridden, because Latin inflectional noun-endings typically reveal the grammatical number of the noun.

pos -- (part of speech) 1=adj 2=adv 3=conj 4=interj 5=noun 6=prep 7=pron 8=verb

tdo -- time-of-direct-object for parser module.

tio -- time-of-indirect-object for parser module.

toc -- time-of-old-concept -- in the Mens Latina Latin AI, is obtained from the index "i" in the AudRecog() auditory recognition module so that other modules in the AI Mind may zero in not on the auditory engram of the word but on the simultaneous conceptual node for the recognized word, in order to harvest or fetch the values in the flag-panel of quasi-neuronal associative tags connecting the one recalled node of the concept with other concepts.

tsj -- time-of-subject for parsing.

tvb -- time-of-verb for parsing.

ux1 -- -- unknown Latin case potentially nominative.

ux4 -- -- unknown Latin case potentially accusative.

uy1 -- -- unknown Latin case potentially nominative.

uy4 -- -- unknown Latin case potentially accusative.

6. Troubleshooting and Debugging

6.1.a. Symptom: (Something goes wrong.)
6.1.b. Solution: (AI Mind Maintainer devises solution.)

7. Future Development of LaParser

Roadmap to Artificial Intelligence

8. Acknowledgements

Mitomino kindly and helpfully wrote a response on the Latin StackExchange.

Draconis kindly and helpfully wrote a response on the Latin StackExchange.

9. Resources for the LaParser Mind-Module

Concursus Omnium Bonorum -- list of Latin-language weblogs

Mens Latina Programming Journal

10. Spread the News on TikTok and Other Venues

Are you on TikTok? Are you eager to be a ThoughtLeader and Influencer?
Create a TikTok video in the following easy steps.

I. Capture a screenshot of
for the background of your viral TikTok video.

II. In a corner of the screenshot show yourself talking about the LaParser module.

III. Upload the video to TikTok with a caption including all-important hash-tags which will force governments and corporations to evaluate your work because of FOMO -- Fear Of Missing Out:
#AI #ИИ #brain #мозг #ArtificialIntelligence #ИскусственныйИнтеллект #consciousness #сознание #Dushka #Душка #psychology #психология #subconscious #подсознание
#AGI #AiMind #Alexa #ChatAGI #chatbot #ChatGPT #cognition #cyborg #Eureka #evolution #FOMO #FreeWill #futurism #GOFAI #HAL #immortality #JAIC #JavaScript #linguistics #metempsychosis #Mentifex #mindmaker #mindgrid #ML #neuroscience #NLP #NLU #OpenAI #OpenCog #philosophy #robotics #Singularity #Siri #Skynet #StrongAI #transhumanism #Turing #TuringTest #volition

A sample video is at

11. AiTree of Artificial Intelligence in Latin Language

MainLoop of Artificial Intelligence in Latin Language
PraeScium mindboot sequence of innate Latin vocabulary
KbLoad -- load the Knowledge Base and initial Latin vocabulary
ReJuvenate -- recycle memory space
Sensorium -- sensory input with associative tagging to concepts
AudInput -- auditory input
AudListen -- listen for input
AudMem -- Latin auditory memory
AudRecog -- auditory recognition
OldConcept -- recognize a known old concept
LaParser -- determine meaning from inflectional endings
InStantiate -- create new instance of concept
NewConcept -- create a new concept
LaParser -- determine part of speech
InStantiate -- create instance of new concept
GusRecog -- sense of taste
OlfRecog -- sense of smell
TacRecog -- sense of touch
VisRecog -- visual recognition by feature extraction
Volition -- free will resulting from emotion and thought
Emotion -- as influence on thought and free will
LaThink -- for thinking in Latin with concepts
Imperative -- mood for issuing commands
Indicative -- mood for declarative statements
cf. EnNounPhrase or cf. EnPronoun -- for subject of thought
cf. EnAdjective -- insert an adjective
LaNounGen -- generate a Latin noun-form
cf. EnVerbPhrase -- think with a verb
cf.EnAdverb -- modifies a verb
cf.EnAdverb -- modifies another adverb
cf. EnVerbGen -- generates missing Latin verb-forms
AudBuffer -- stores Latin phonemes
OutBuffer -- right-justifies Latin verb-form
Speech -- output to screen or loudspeaker
cf. EnNounPhrase or VisRecog -- direct-object noun
cf. EnPrep -- LaPrep uses a Latin preposition
cf. EnNounPhrase or cf. EnPronoun -- object of preposition
cf. ConJoin -- insert a conjunction if two thoughts are active
Subjunctive -- mood for conditional phrases
PsiDecay -- lowers conceptual activation
SpreadAct -- spreading activation from concept to concept
MetEmPsychosis -- AI soul travel from webserver to webserver
MindMeld -- merge two AI Minds by sharing memories
Motorium (MotorOutPut) -- for AI-Minded robots
Consciousness -- emergent property, not a module

Nota Bene: This webpage is subject to change without notice. Any Netizen may copy, host or monetize this webpage to earn a stream of income by means of an affiliate program where the links to Amazon or other booksellers have code embedded which generates a payment to the person whose link brings a paying customer to the website of the bookseller.

This page was created by an independent scholar in artificial intelligence who created the following True AI Minds with sentience and with limited consciousness. -- MindForth Robot AI in English. -- Forth Robot AI in German. -- Robot AI thinks in English and in Russian. -- JavaScript Robot AI Mind thinks in English. -- JavaScript Robot AI Mind thinks in Latin. -- JavaScript Robot AI Mind thinks in Russian.

The following books describe the free, open-source True AI Minds.

AI4U --

AI4U (paperback) --

AI4U (hardbound) --

The Art of the Meme (Kindle eBook) --

Artificial Intelligence in Ancient Latin (paperback) --

Artificial Intelligence in Ancient Latin (Kindle eBook) --

Artificial Intelligence in German (Kindle eBook) --

InFerence at Amazon USA (Kindle eBook) --

563 Mentifex Autograph Postcards were mailed in 2022 primarily to autograph collector customers at used bookstores to press the issue of whether or not the Mentifex oeuvre and therefore the autograph is valuable. These artwork-postcards with collectible stamps may be bought and sold at various on-line venues.

See AI 101 AI 102 AI 103 year-long community college course curriculum for AI in English.
See Classics Course in Latin AI for Community Colleges or Universities.
See College Course in Russian AI for Community Colleges or Universities.
Collect one signed Mentifex Autograph Postcard from 563 in circulation.
See Attn: Autograph Collectors about collecting Mentifex Autograph Postcards.

Return to top; or to
Image of AI4U found at a book store
The collectible AI4U book belongs in every AI Library as an early main publication of Mentifex AI.

-- -- Mentifex weblog -- TikTok

Arno Welzel

Jul 30, 2023, 3:48:10 AMJul 30
A.T. Murray, 2023-07-30 00:13:

> LaParser Latin Parsing Module of the
> Mens Latina Artificial Intelligence in Latin Language


Arno Welzel

Reply all
Reply to author
0 new messages