Episodic Auto-translation with Semantic/ Syntactic/Episodic (Words/Grammar/ Episodic) Model: Word Character Division, Frequent Binding and Inter-sentential Congruency | Chapter 9 | An Overview of Literature, Language and Education Research Vol. 1

From 2015 to now, it has been formulated by Cai the semantic/syntactic/ episodic(words/grammar/episodic) model of native language, corresponding to the declarative remote memory, procedural memory and declarative immediate/recent episodic memory, respectively, with the N400-related episodic congruency of word-meaning necessary to determine one congruent meaning for some polymeaning words in sentence and paragraph. Herein, it is aimed to apply this model to machine translation, as follows: (a) From the words/grammar/episodic components, it is derived three principles for machine translation, as word/phrase dictionary, procedural grammar input/output, and episodic congruency of polymeaning-words, respectively. (b) It is classified the episodic congruency of word-meaning in sentences and paragraphs into three types as stepwise congruent comprehension by gamma increment; frequent word binding from phrases, idioms, etc; and inter-sentential association of polymeaning-word in a simple sentence with other sentences in a paragraph. (c) For stepwise congruent comprehension, it is classified the pronouns/nouns into the categories of behaviorally alive human/animal and naturally changing matter, and further classified such nouns/pronouns into the zoological/organizational/ range/relation/logic/physical/categorical characters, and the sententially stepwise following verbs into affective/behavioral/range/relation/ logic/characteristic/ changing characters, the adjectives/adverbs into affective/behavioral/ range/ relation/logic/characteristic/changing/spatial/temporal characters, restricting their meanings by congruency of word category of characters in sentence. (d) For frequent word bindings, it is suggested to make a databank to collect and use them. (e) Finally, it is suggested to further adopt statistical translation by paragraph to determine the polymeaning-word in simple sentences. In conclusion, it is feasible to progress significantly with such applications to machine translation.

 

Author(s)details:-

 

Zi-Jian Cai
Cai Fortune TriL Consulting No. 129, Building 6, Room 404, North Dongwu Road, Suzhou City, Jiangsu Province, 215128 PR, China.

 

Please See the book here :- https://doi.org/10.9734/bpi/aoller/v1/703