首页|Findings in the Area of Machine Translation Reported from Shandong Management University (Automatic Recognition of Machine English Translation Errors Using Fuzzy Set Algorithm)
Findings in the Area of Machine Translation Reported from Shandong Management University (Automatic Recognition of Machine English Translation Errors Using Fuzzy Set Algorithm)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Springer Nature
Investigators discuss new findings in Machine Translation. According to news reporting from Shandong, People's Republic of China, by NewsRx editors, the research stated, "Fuzzy sets demonstrate remarkable efficacy in addressing a wide range of challenges in real-world domains, surpassing the capabilities of traditional approaches. These disciplines include data analysis, machine learning, decision theory, data mining, recognition tasks, intelligence, and hybrid systems." The news correspondents obtained a quote from the research from Shandong Management University, "As a result, the application of fuzzy sets extends to diverse areas, such as robotics, intelligent systems, medical and satellite systems, decision-making in consumer electronics, information processing, pattern recognition, and optimization. Nowadays, one of the applications, language recognition, is a particular issue-precisely, the error frequency in the machine language translator. The frequency of errors in simple machine English translation is increasing day by day. With modern information technology's continuous evolution and development, simple machine translation has yet to meet people's normal needs. This research paper presents a novel machine translation framework founded on automatic error detection. In the realm of machine translation, effectively incorporating user feedback alongside linguistic knowledge remains a challenge. To address this complexity, the study advocates employing a machine learning technique, specifically the fuzzy set algorithm, to extract valuable insights. These insights are instrumental in refining machine-generated translations into more standardized, accurate outputs. The application of this knowledge to other machine translations aims to rectify common errors, ultimately enhancing the overall usability of machine translation systems. Through iterative experiments, the study expanded its set of translation rules, extracting 50 and 100 rules by iteratively adjusting translations through addition, deletion, and modification. Interestingly, the research found that an excessive number of iterations did not necessarily lead to improved translation quality; instead, stabilization occurred after rule sequences. Additionally, the study delved into automatic error identification in machine-generated English translations, introducing automatic post-editing technology to significantly enhance translation quality."
ShandongPeople's Republic of ChinaAsiaAlgorithmsCyborgsEmerging TechnologiesFuzzy LogicMachine LearningMachine TranslationTechnologyShandong Management University
NETL
NSTL
Springer Nature
