Verification of Consistency Between Process Models, Object Life Cycles, and Context-Dependent Semantic Specifications

Hoch, Ralph; Luckeneder, Christoph; Popp, Roman; Kaindl, Hermann

doi:10.1109/tse.2021.3110191

Cited by 2 publications

(1 citation statement)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Despite this, a very large number of fixed-length sequences of consecutive words will still be produced, so the sequences of consecutive words containing stop words in the beginning or end position are deleted. English translation stops words for papers, adverbs, conjunctions, and other words without practical significance; these words cannot express the meaning of the statement but only play the role of the successive and transitive, so delete these phrases [16].…”

Section: A Keyword Extraction Algorithm For English Translationmentioning

confidence: 99%

A Deep Learning-based Method for Determining Semantic Similarity of English Translation Keywords

Zhili,

Qian

2024

ijacsa

View full text Add to dashboard Cite

In the English translation task, the semantics of context play an important role in correctly understanding the subtle differences between keywords. The bidirectional LSTM includes a positive LSTM and a reverse LSTM. When processing sequence data, you can consider the information of the preceding and following text at the same time. Therefore, to capture the subtle semantic differences between English translation keywords and accurately evaluate their similarity, a new semantic similarity determination method for English translation keywords is studied with the bidirectional LSTM neural network in deep learning as the main algorithm. This method introduces an English translation keyword extraction algorithm based on word cooccurrence and uses the co-occurrence relationship between words to identify and extract keywords in English translation. The extracted keywords are input into the bidirectional LSTM neural network keyword semantic similarity judgment model based on deep learning, and the weight of the bidirectional LSTM neural network is set by using the sparrow search algorithm to optimize. After the bidirectional LSTM neural network is trained, the information on keyword word vectors is captured, and the similarity between keyword word vectors is evaluated. The experimental results show that the sentence similarity calculated by the proposed method for English translation is very close to the result of professional manual scoring. The Spearman rank correlation coefficient of the semantic similarity determination result is 1, and the determination result is accurate.

show abstract

Section: A Keyword Extraction Algorithm For English Translationmentioning

confidence: 99%