Article Preview
Top1 Introduction
Any language in the world contains words that can be classified into name, verb, preposition, adverb, pronoun, etc. that called part of speech. From a set of words composed a sentence, this sentence has grammar in the way it is written. The grammar for writing a sentence varies from one language to another. These grammars are called context free grammar. To recognize any sentence, we need to know the grammar of this language.
English language, like any other language, contains parts of speech. Also, most sentences contain a verb. A verb depicts an event or of a case is. The nature and time of the event are signed by tense (Tiwari, 2008). The index of tenses varies according to the nature of language. In English language, tense is regularly represented through verb intonations and therefore can be simply identified (Chomsky, 1957).
In linguistics, morphology is the analysis of phrases (words), how they are shaped, and their association to other words in the similar language. In English language, morphology is a portion of English grammar which studies the construction of English words, their components and meanings and the shape of the words. The English morphology studies the root words, stems, inflections prefixes, suffixes, and phonemes. Several Text Analytics applications require morphological analyzers to complete their tasks where they are considered as preprocessors for text analysis(Narejo & Mahar, 2016).
Syntax analysis has been dependent in this paper, which is an essential region of research in computational linguistics. Therefore, recognizing the syntactic construction is integral in figuring out the connotation of the sentence. The determination is performed by a process identified as parsing. Parsing dealing with the syntactic shape of a sentence (Martin, 2000) (Massa Cereda et al., 2018).
There are works close to the subject of this research, such as: in (Ge et al., 2015) worked on predicting tense in chinse conversation. Since verbs in the Chinese language did not have clear grammatical or lexical forms to show tense. Also, tense information is often implicitly hidden outside of the target sentence. Therefore, they used linguistic resources to propose a set of novel sentence-level (local) features and a new hypothesis of “One tense per scene” to incorporate scene-level (global) evidence. Their experimental results proved the power of this approach. Also, in (Lee, 2011) worked on presenting a technique to predict verb tenses using a statistical model taught on conditional random fields. The result of this model outperforms the baseline. While in (Gong et al., 2012) worked on mapping tenses from one language(source) to another language(target) in machine translation. The source language acks over tense markers while the target language recognized tense information. Therefore, they proposed a classifier based tense model to save the primary tense in the destination side compatible with the one in the source side. In several NLP applications, such as event summarization and extraction, tense considers as a key element in presenting the temporal order of events (Karthikeyan et al., 2019). From verb tenses prediction, one can define the temporal ordering of events in a report and additionally can conclude temporal relations inside the sentence(Lee, 2011) (Ferreira, 2017). Therefore, tense prediction is important in simultaneous machine translation and sentence processing application (Gong et al., 2012) (Sultana et al., 2019) (Hussein et al., 2014) (Kumar et al., 2021).
In this paper, our goal is to design a model to automatically recognize sentences’ tenses in English articles/stories by implementing morphological and syntax analysis, taking into account local syntactic, involving temporal features.
The paper is organized as follows as: section 1 is about Introduction. section 2 is about the Proposed tense recognition model. section 3 is about experimental result of proposed work. Section 4 is about conclusion and future work.