The Role of Error Type and Working Memory in Written Corrective Feedback Effectiveness on First-Language Self Error-Correction

This study examined the role of error-type and working memory (WM) in the effectiveness of direct-metalinguistic and indirect written corrective feedback (WCF) on self error-correction in first-language writing. Fifty-one French first-year psychology students volunteered to participate in the experiment. They carried out a first-language error-correction task after receiving WCF on typographical, orthographic, grammatical, and semantic errors. Results indicated that error-type affected the efficacy of WCF. In both groups, typographical error-correction was performed better than the others; orthographic and grammatical error-correction were not different, but both were corrected more frequently than semantic errors. Between-group comparisons showed no difference between the two groups in correcting typographical, orthographic, and grammatical errors, while semantic error-correction was performed significantly better for the direct group. Results revealed that WM was not involved in correcting typographical, orthographic, and grammatical errors in both groups. It did, however, predict semantic error-correction only in response to direct-metalinguistic WCF. In addition, the processing component of WM was predictive of semantic error-correction in the direct WCF group. These findings suggest that error-type mediates the effectiveness of WCF on written error-correction at the monitoring stage of writing, while WM does not associate with all WCF types efficacy at this stage.

An Analysis of Lexical and Syntactical Errors in English Abstract Made by Students of the Accounting Department Polmed

The main purpose of the study is to find out the kinds of lexical and syntactical errors in writing the student’s paper of English Abstract. The data collecting technique used the field research method from 25 student’s papers of English abstract of Accounting Department Politeknik Negeri Medan. The method used in the research is a qualitative approach. The findings of the research revealed that there were so many errors found in the student’s papers of English abstract. There were two kinds of lexical errors relate to Semantic Error in Lexis found in the student’s papers of English abstract which are Confusion of Sense Relations and Collocational Errors. The number of errors in Confusion of Sense was 20 errors with percentage 62.50 % while the collocational errors were also found as many as 12 errors that constituted 37,50% of the total errors collected in the data analysis process. The errors in the syntactical level are also found in the student’s paper of English abstract based on surface structure taxonomy. There were five kinds of errors found in the analysis process. The most frequent errors made by students are misformation 62 (40,79%) of error followed by omission error reach 54 (35,53%) of errors. The addition errors reach 22 (14,47%) of error. then, misordering errors are found 13 (8,55%) of errors. finally, the blend is only found 1 (0,66%) of errors in student’s papers of English abstract in Accounting Department Politeknik Negeri Medan.

THE ANALYSIS OF ERRORS MADE BY THE THIRD YEAR STUDENTS OF ENGLISH DEPARTMENT IN TRANSLATING NARRATIVE TEXT

This study aims at finding out the errors made by the third year English Department students of Universitas Negeri Padang in translating a narrative text from English into bahasa Indonesia. This study used a descriptive method. The data were the translation of the narrative text entitled Snow White done by the third year English Department students of Universitas Negeri Padang. The data were collected using a test and were analyzed using the error analysis. The results of the study indicate semantic error is the error frequently committed by the students in their translation with the frequency of 72 times (59%). Pragmatic error (29%) is the second most frequent error, while morphological error and syntactical error are seldom found in the students’ translation. In conclusion, the errors made by the students are mainly motivated by their tendency to follow the meaning of the dictionary (word for word translation).

Applying a Hybrid Sequential Model to Chinese Sentence Correction

In recent years, Chinese has become one of the most popular languages globally. The demand for automatic Chinese sentence correction has gradually increased. This research can be adopted to Chinese language learning to reduce the cost of learning and feedback time, and help writers check for wrong words. The traditional way to do Chinese sentence correction is to check if the word exists in the predefined dictionary. However, this kind of method cannot deal with semantic error. As deep learning becomes popular, an artificial neural network can be applied to understand the sentence’s context to correct the semantic error. However, there are still many issues that need to be discussed. For example, the accuracy and the computation time required to correct a sentence are still lacking, so maybe it is still not the time to adopt the deep learning based Chinese sentence correction system to large-scale commercial applications. Our goal is to obtain a model with better accuracy and computation time. Combining recurrent neural network and Bidirectional Encoder Representations from Transformers (BERT), a recently popular model, known for its high performance and slow inference speed, we introduce a hybrid model which can be applied to Chinese sentence correction, improving the accuracy and also the inference speed. Among the results, BERT-GRU has obtained the highest BLEU Score in all experiments. The inference speed of the transformer-based original model can be improved by 1131% in beam search decoding in the 128-word experiment, and greedy decoding can also be improved by 452%. The longer the sequence, the larger the improvement.

Mutalyzer 2: Next Generation HGVS Nomenclature Checker

Unambiguous variant descriptions are of utmost importance in clinical genetic diagnostics, scientific literature, and genetic databases. The Human Genome Variation Society (HGVS) publishes a comprehensive set of guidelines on how variants should be correctly and unambiguously described. We present the implementation of the Mutalyzer 2 tool suite, designed to automatically apply the HGVS guidelines so users do not have to deal with the HGVS intricacies explicitly to check and correct their variant descriptions. Mutalyzer is profusely used by the community, having processed over 133 million descriptions since its launch. Over a five year period, Mutalyzer reported a correct input in approximately 50% of cases. In 41% of the cases either a syntactic or semantic error was identified and for approximately 7% of cases, Mutalyzer was able to automatically correct the description.

The Precision and Accuracy of English Verbs Choice in Recount Text Written by the Students of CEC Denpasar: An Analysis of Metalanguage

<p>This writing entitled “The Precision and Accuracy of English Verbs Choice in Recount Text Written by The Students of CEC Denpasar: An Analysis of Metalanguage” was conducted in order to be able to find out the percentage of precision and accuracy of English verbs choice in recount text written by the students of CEC Denpasar, analyze the lexico-semantic error, and describe the factors that cause this error. The data in this study were taken directly from recount texts written by the students of CEC Denpasar, interview, and questionnaire. Those data were collected by using a test, interview, and questionnaire method. The theory applied in this writing is Metalanguage from NSM theory proposed by Wierzbicka (1996). However, in describing the error, the analysis was done through configuration only by the qualitative method. The result shows that out of twenty-six verbs, twelve chosen verbs were less precise. Through Metalanguage, those errors can be described configurationally. Moreover, the difference between the error and the recommendation verbs can be explained, therefore, the difference can be shown. There were two factors that cause this error. They are external (teacher, method, strategy, technique used in teaching) and internal (motivation, potency, competence, and learners’ language background) factors.</p>

Location-Based End-to-End Speech Recognition with Multiple Language Models

End-to-End deep learning approaches for Automatic Speech Recognition (ASR) has been a new trend. In those approaches, starting active in many areas, language model can be considered as an important and effective method for semantic error correction. Many existing systems use one language model. In this paper, however, multiple language models (LMs) are applied into decoding. One LM is used for selecting appropriate answers and others, considering both context and grammar, for further decision. Experiment on a general location-based dataset show the effectiveness of our method.