Incorrect Word Research Articles

Monotonic Gaussian regularization of attention for robust automatic speech recognition

The Attention-based Encoder–Decoder (AED) models are one of the most popular models for Automatic Speech Recognition (ASR). However, instability can occur in AED with problems such as incorrect insertions or word repetitions due to the violation of the inherent monotonic alignment property. To address these problems, we propose a monotonic Gaussian regularization method to guide the attention training, where the guiding map is depicted as a sequence of Gaussian distributions with monotonically moving centers. Experiments show our method reduces the insertion error rate by a relative 7% on the HKUST dataset, relative 20% and 16% on two large industrial datasets, and a relative 21% on an out-of-domain test set. The overall Character Error Rates (CERs) are all reduced at the same time, indicating that the model’s recognition ability is well maintained. Therefore, our proposed method improves model performance by enhancing monotonic alignment, and provides better robustness.

The Error Analysis of Google Translate and Bing Translator in Translating Indonesian Folklore


 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 The objective of this research is to find out the lexical errors made by Google Translate and Bing Translator in translating Indonesian folklore "Princess Tandampalik" and "Sigarlaki and Limbat." The research applied the qualitative method. The data were analyzed using hybrid taxonomy of error analysis from Vilar, et al. The results of this research show that Google Translate made 103 errors in total which consist of 12 missing words, 19 errors in word order, 64 incorrect words, and 8 unknown words. Meanwhile, Bing Translator made 95 errors which consist of 5 missing words, 1 error in word order, 88 incorrect words, and 1 unknown word. Incorrect word is the most frequent error found in the translation resulted from Google Translate and Bing Translator with a total of 152 errors. The incorrect words mainly occurred in the translation of adjectives and adverbs in which Google Translate and Bing Translator mostly translated them into noun form. Thus, it can be concluded that both machine translators' performances are not different because they have their advantages and disadvantages. 
 

Synergic Concepts, Lexical Idiosyncrasies, and Lexical Complexities in Bilingual Students’ Translated Texts as Efforts to Resolve Conceptual Inequivalences

The purpose of this study is to draw on the conceptual blending hypothesis from the socio-cognitive approach to investigate the conceptually equivalent translation written in L2—English—of bilingual students via two tasks of translating and defining individual words and translating texts from L1 to L2. Next, the study demonstrates how translation abilities that vary amongst groups can affect students’ lexical density, lexical diversity, lexical sophistication, and lexical idiosyncrasies in translated text. The translating process in bilinguals could be interpreted via the lens of the conceptual blending hypothesis and dueling contexts framework to demonstrate that bi/multilingual students do not differ from monolingual ones pertaining to cognitive or linguistic abilities. Rather, the distinctive difference between bilingual and monolingual language users is bilingual speakers’ abilities of the third competence of formulating a synergism across word concepts and utilizing a bidirectional translation between two languages. When a word in L2 is acquired, there is a conceptual blending between the new conceptual information, encoded after each time the L2 word is used in an L2 socio-cultural context and the existing socio-cultural conceptual information in L1. The new concept created after the blending is called a synergic concept. If the synergic is not well developed, the language user selects incorrect or inappropriate words in a context, resulting in lexical idiosyncrasies. Data gathered from 30 English–Chinese bilingual university students in a transnational program in sociology were collected and compared against 15 monolingual American students. The preliminary findings are as follows: (1) regardless of the location of where the English (L2) socio-cultural meaning conceptualization mainly takes place (in China or the U.S.A.), English–Chinese bilingual language users demonstrated a significant difference in connotative meaning knowledge of noun word concepts and idiomatic concepts, compared with English native speakers; (2) the synergic concepts were detected in all experimental concepts and demonstrated the conceptual blending to a varying degree that affects their translating process and its outcomes: the domineering L1 socio-cultural concept, the well-blended L1 and L2 socio-cultural concept that results in a “third culture”, and the assimilating L2 socio-cultural concept; (3) the synergistic blending of two socio-cultural loads embedded in lexical concepts detected in the bilingual students in the U.S.A. was more robust than those in China, resulting in significantly fewer sophisticated words and lexical idiosyncrasies in their English translated essays. The study sheds new light on understanding the dynamism in bilingualism via translation tasks to indicate bilingual learners’ lexical development. Implications for using translation tasks and analysis of word concepts across languages to support bi/multilingual students in language and academic learning are discussed.

IMITATION TECHNIQUE IN LEARNING ENGLISH AT ENGLISH EDUCATION DEPARTMENT UNIVERSITAS NEGERI MANADO

English as a foreign language is something interest to learn even though it is not easy to learn. In learning English many teachers were try some media, method, strategy, and technique. The aim of this research is to describe how the use of imitation technique is effective in learning English. In this research the researcher used imitation technique in learning English to improve students’ pronunciation and listening skill. By used video the students imitate what are the speaker say, before that they must listen carefully what speaker say. In this research consists of 20 students in third semester and used three videos. This research applied in three treatments. The result showed the student receive better improvement in learning English. The findings in the analysis it showed the improvement of students’ pronunciation is increased from 22.85 to 0.95 incorrect words and the improvement of listening is increased from 28.2 to 78.7, and it concluded that imitation technique is effective technique in learning English. When the words pronounced correctly it was help to recognize the words in case listening skill. In COVID-19 pandemic it is possible to use media such as video youtube in teaching and learning process even though technology could not replace the teacher position but technology will be distributor the knowledge from teacher to student.

PENGEMBANGAN INSTRUMEN TES KETERAMPILAN MENYIMAK UNTUK SISWA KELAS IX SEKOLAH MENENGAH PERTAMA

Tulisan ini bertujuan mendeskripsikan prosedur pengembangan dan bentuk instrumen tes keterampilan menyimak untuk siswa kelas IX Sekolah Menengah Pertama yang dihasilkan. Penelitian ini berbasis daring menggunakan google formulir. Pengembangan menggunakan model ADDIE. Pengembangan dilakukan beberapa tahap, yaitu: (1) analisis kebutuhan (Analysis), (2) desain produk (Design), (3) pengembangan produk awal (Development), (4) uji coba produk awal (Implementation), (5) revisi produk awal (Evaluation), (6) uji coba produk revisi, (7) revisi produk revisi. Uji coba dilakukan kepada siswa kelas IX SMP/MTs dengan tekni accidental sampling. Pengembangan ini menghasilkan tes berbasis google form yang terdiri dati biodata untuk peserta tes, rekaman cerpen sebagai bahan simakan berdurasi 11 menit , 25 soal (22 pilihan ganda dan 3 jawaban singkat) lengkap dengan kunci jawaban berjumlah 951 kata, memiliki skor yang berbeda pada setiap soal, dan rubrik penilaian tes keterampilan menyimak. Uji coba produk awal diperoleh responden sebanyak 20 orang. Revisi yang peneliti lakukan, yaitu (1) merevisi soal dan petunjuk pengerjaan soal, (2) mengganti beberapa komponen penilaian yang kurang baik, dan (3) mengganti beberapa pilihan kata yang kurang tepat. Revisi yang peneliti lakukan, yaitu (1) merevisi durasi instrumen tes keterampilan menyimak, (2) merevisi soal dan petunjuk pengerjaan soal, dan (3) mengganti beberapa pilihan kata yang kurang tepat.

Predictive Sentence Context Reduces Listening Effort in Older Adults With and Without Hearing Loss and With High and Low Working Memory Capacity.

Listening effort is needed to understand speech that is degraded by hearing loss, a noisy environment, or both. This in turn reduces cognitive spare capacity, the amount of cognitive resources available for allocation to concurrent tasks. Predictive sentence context enables older listeners to perceive speech more accurately, but how does contextual information affect older adults' listening effort? The current study examines the impacts of sentence context and cognitive (memory) load on sequential dual-task behavioral performance in older adults. To assess whether effects of context and memory load differ as a function of older listeners' hearing status, baseline working memory capacity, or both, effects were compared across separate groups of participants with and without hearing loss and with high and low working memory capacity. Participants were older adults (age 60-84 years; n = 63) who passed a screen for cognitive impairment. A median split classified participants into groups with high and low working memory capacity. On each trial, participants listened to spoken sentences in noise and reported sentence-final words that were either predictable or unpredictable based on sentence context, and also recalled short (low-load) or long (high-load) sequences of digits that were presented visually before each spoken sentence. Speech intelligibility was quantified as word identification accuracy, and measures of listening effort included digit recall accuracy, and response time to words and digits. Correlations of context benefit in each dependent measure with working memory and vocabulary were also examined. Across all participant groups, accuracy and response time for both word identification and digit recall were facilitated by predictive context, indicating that in addition to an improvement in intelligibility, listening effort was also reduced when sentence-final words were predictable. Effects of predictability on all listening effort measures were observed whether or not trials with an incorrect word identification response were excluded, indicating that the effects of predictability on listening effort did not depend on speech intelligibility. In addition, although cognitive load did not affect word identification accuracy, response time for word identification and digit recall, as well as accuracy for digit recall, were impaired under the high-load condition, indicating that cognitive load reduced the amount of cognitive resources available for speech processing. Context benefit in speech intelligibility was positively correlated with vocabulary. However, context benefit was not related to working memory capacity. Predictive sentence context reduces listening effort in cognitively healthy older adults resulting in greater cognitive spare capacity available for other mental tasks, irrespective of the presence or absence of hearing loss and baseline working memory capacity.

The Role of Computers in the Development of Odia Language

Abstract: Any language computer can be useful if the above two requirements are met. But we can only accept such behavior as a modern writing system. In other words, in the past we used to write with palm leaves and pencils, then we used to write on paper, then we used typewriters and typewriters. Now we can write via computer. This is the best way to write and print. This requirement is available in almost all languages, this requirement has been met by the Odia language since the 180s, which means that this year we have been able to type the letters of the Odia language on the computer and print it on the printer. This is the first step in using a computer language. The second step in making a language useful to a computer is: understanding the computer. Computers understand a language through a variety of programs. This requires connecting the operating system to a new language. It is imperative that the language be integrated into the Unicode system. Unicode has provided a code for the scripts of the world's most computer-friendly languages. This code can be understood by the operating system. As a result, it is possible to Unicode the computer by naming that language, sorting it from scratch, searching for a file or folder named in that language, deleting any incorrect word koji, and so on. Fortunately, the Odia language scripts were included in the Unicode in 2006. Odia is the third largest Indian language in this regard. Keywords: typewriters, computer, languages, Unicode, folder, deleting, scripts, Odia, Indian language,

During Lipreading Training With Sentence Stimuli, Feedback Controls Learning and Generalization to Audiovisual Speech in Noise

This study investigated the effects of external feedback on perceptual learning of visual speech during lipreading training with sentence stimuli. The goal was to improve visual-only (VO) speech recognition and increase accuracy of audiovisual (AV) speech recognition in noise. The rationale was that spoken word recognition depends on the accuracy of sublexical (phonemic/phonetic) speech perception; effective feedback during training must support sublexical perceptual learning. Normal-hearing (NH) adults were assigned to one of three types of feedback: Sentence feedback was the entire sentence printed after responding to the stimulus. Word feedback was the correct response words and perceptually near but incorrect response words. Consonant feedback was correct response words and consonants in incorrect but perceptually near response words. Six training sessions were given. Pre- and posttraining testing included an untrained control group. Test stimuli were disyllable nonsense words for forced-choice consonant identification, and isolated words and sentences for open-set identification. Words and sentences were VO, AV, and audio-only (AO) with the audio in speech-shaped noise. Lipreading accuracy increased during training. Pre- and posttraining tests of consonant identification showed no improvement beyond test-retest increases obtained by untrained controls. Isolated word recognition with a talker not seen during training showed that the control group improved more than the sentence group. Tests of untrained sentences showed that the consonant group significantly improved in all of the stimulus conditions (VO, AO, and AV). Its mean words correct scores increased by 9.2 percentage points for VO, 3.4 percentage points for AO, and 9.8 percentage points for AV stimuli. Consonant feedback during training with sentences stimuli significantly increased perceptual learning. The training generalized to untrained VO, AO, and AV sentence stimuli. Lipreading training has potential to significantly improve adults' face-to-face communication in noisy settings in which the talker can be seen.

Attention virtual adversarial based semi‐supervised question generation

AbstractQuestion generation (QG) refers to the automatic generation of questions based on the given passages and answers, and has a wide range of application scenarios in human–computer interaction, education, medical and other fields. However, for Chinese QG, due to the lack of word separation in the writing rules of Chinese text, many methods are not suitable for it, and the generated results have incorrect word order and invalid expressions. In addition, traditional models only use labeled data, but it is difficult and expensive to obtain data labels. In order to solve this problem, this article proposes a semi‐supervised QG model termed virtual stroke‐aware copy network (VSAC Net). It is based on virtual adversarial training and can be used for Chinese QG tasks with few labeled samples. The VSAC Net model combines word embedding virtual counter disturbance and attention virtual counter disturbance, the fitting of the input layer and the attention layer is taken into account, and reduces the overfitting of the model. According to Dureader dataset, a small sample QG dataset is constructed, and the VSAC Net is used for solving. The results show that the proposed model can achieve a better generation effect on small sample datasets.

Lexical Errors Produced by Google Translate in Translating “Putri Serindang Bulan” to English Language

The purpose of this study was to analyze the lexical errors which was produced by google translate in translating “Putri Serindang Bulan” short story to English language as the target language. This study was conducted by using descriptive design with qualitatitive approach, which was done with consideration that the purpose of this research to analyze the lexical errors which produced by google translate in translating “Putri Serindang Bulan” short story to English language as the target language. This study was focused on describing the types of lexical errors produced by google translation in translating “Putri Serindang Bulan” short story to English language. In this research, the writer became the main instrument. This study obtained the data by reading “Putri Serindang Bulan” short story to understand the information of the short story, translating the “Putri Serindang Bulan” short story into English language use google translate, and then comparing the source text with the translation result to find out the translation accuracy and make the terms that were considered as inaccurate translation. The result showed that many errors were produced by google translate in translating the short story, the most frequent error found were incorrect word and missing word. In conclusion, this study suggest that the human translators must revise the translation result from machine translation.

AN ERROR ANALYSIS OF SEMANTIC LEVEL IN RIAU NEWSPAPER ARTICLES

Language errors often occur when communicating, both oral communication and written communication. This happens because of the lack of understanding of the communicator in delivering messages to the communicant or the author's lack of understanding in providing information to his readers. In addition, language errors also often occur due to spoken language factors, so that when applied in written language, there tend to be errors. Language errors at the semantic level still often occur in written language, especially articles in the Riau newspaper. These errors can be in the form of errors due to confused pairs, errors due to incorrect choice of words, and errors due to cognate pairs. This study uses a prescriptive content analysis method with the data source from the newspaper Koran Riau. Data is collected by means of documentation with data collection techniques including, collecting error samples, identifying errors, explaining errors, classifying errors, and evaluating or correcting errors. Based on the results and discussion, there are two errors in the language of the semantic level in the Riau Koran newspaper article. First, the fault of a confused partner. Second, errors due to incorrect word choices. However, errors due to cognate pairs were not found.

Correction to: Fingerspitzengefühl

Risk aversion kicks in: the less risk averse the insured, the larger the deviation from full insurance, all else being equal. The original version of this paper was inadvertently published with an incorrect word in one sentence on page 255. The original article has been corrected. We apologise for any inconvenience caused to our readers.

Radiology Dictation Errors with COVID-19 Protective Equipment: Does Wearing a Surgical Mask Increase the Dictation Error Rate?

Our aim was to determine the effect of wearing a surgical mask on the number and type of dictation errors in unedited radiology reports. IRB review was waived for this prospective matched-pairs study in which no patient data was used. Model radiology reports (n = 40) simulated those typical for an academic medical center. Six randomized radiologists dictated using speech-recognition software with and without a surgical mask. Dictations were compared to model reports and errors were classified according to type and severity. A statistical model was used to demonstrate that error rates for all types of errors were greater when masks are worn compared to when they are not (unmasked: 21.7 ± 4.9 errors per 1000 words, masked: 27.1 ± 2.2 errors per 1000 words; adjusted p < 0.0001). A sensitivity analysis was performed, excluding a reader with a large number of errors. The sensitivity analysis found a similar difference in error rates for all types of errors, although significance was attenuated (unmasked: 16.9 ± 1.9 errors per 1000 words, masked: 20.1 ± 2.2 errors per 1000 words; adjusted p = 0.054). We conclude that wearing a mask results in a near-significant increase in the rate of dictation errors in unedited radiology reports created with speech-recognition, although this difference may be accentuated in some groups of radiologists. Additionally, we find that most errors are minor single incorrect words and are unlikely to result in a medically relevant misunderstanding.

Detecting and correcting automatic speech recognition errors with a new model

The purpose of automatic speech recognition (ASR) systems is to recognize speech signals obtained from people and convert them into text so that they can be processed by a computer. Although many ASR applications are versatile and widely used in the real world, they still generate relatively inaccurate results. They tend to generate spelling errors in recognized words, especially in noisy environments, in situations where the vocabulary size is increased, and at times when the input speech is of poor quality. The permanent presence of errors in ASR systems has led to the need to find alternative methods for automatic detection and correction of such errors. In this study, the basic principles of ASR evaluation are first summarized, and then a new approach based on the suggestion of an alternative hypothesis is proposed for the detection and correction of these errors generated by ASR systems. The proposed method involves a series of processes such as identifying incorrect words, selecting the ones that can be corrected, and identifying candidate words to replace these words. As a result of the tests carried out by creating different test environments, significant performance improvements for Turkish were achieved and an average of 4.60 % performance improvement was provided.

Differential effects of pressure on social contagion of memory.

Collaboration can support performance on a variety of tasks, but recent projects have indicated that group collaborations can also be associated with memory decrements. For example, when people discuss ideas, any shared inaccurate information can be used by group members to complete subsequent tasks. Across two experiments, we examined whether this social contagion is influenced by performance pressures that regularly emerge during group interactions. In Experiment 1, participants under individual-directed pressure, goal-directed pressure, or control conditions studied word lists before completing a collaborative recall with a confederate partner who occasionally recalled incorrect words. We examined whether partner-produced inaccuracies contaminated participants' memories for the list contents on a subsequent individual recall. Goal-directed pressure, focused on monitoring partner accuracy, facilitated subsequent individual recall, as demonstrated by reduced reproductions of partner-generated inaccuracies. In contrast, individual-directed pressure, focused on appearing competent, resulted in greater use of those inaccuracies. Experiment 2 ruled out that the benefits associated with goal-directed pressure were solely due to warning participants about the possibility of social contagion. These results demonstrate that different instantiations of pressure can help or hinder memory-related performance in collaborative settings. Under the right conditions, pressure can help overcome the effects of exposure to inaccurate information. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Correction to: Diagnostic Performance of Fundus Autofluorescence for Detecting Polypoidal Choroidal Vasculopathy

The author has informed the editor to correct the legend of figure and word in the article as follow: Incorrect legend of figure in the article: Figure 2. Optical coherence tomography of clinical signs of PCV: (A) steep pigment epithelial detachment (sPED); (B) notched pigment epithelial detachment (nPED); (C) double-layer sign (DLS); and (D) hyporeflective lumen, with retinal pigment epithelial detachment. Correct legend of figure: Figure 2. Optical coherence tomography of clinical signs of PCV: (A) steep pigment epithelial detachment (sPED); (B) double-layer sign (DLS); (C) notched pigment epithelial detachment (nPED); and (D) hyporeflective lumen, with retinal pigment epithelial detachment. Incorrect word in the article: What is already known on this topic? FAF is an invasive investigation aimed for RPE evaluation that is used widely in many retinal diseases. Correct word: What is already known on this topic? FAF is a non-invasive investigation aimed for RPE evaluation that is used widely in many retinal diseases.

The press of the South of Ukraine about the struggle of African Americans for rights and freedoms after the abolition of slavery in the United States

The article examines the publications of the Odesa press of the second half of the 19th century on coverage of issues of granting rights and freedoms to African Americans after the abolition of slavery in the United States. The regional press formed the public opinion of the South of Ukraine. The main problems that were raised on the pages of the periodicals were identified. The main areas of reflection of life abroad were assessments of internal, economic and foreign policy development of countries. The main problems that were raised in the materials about the United States on the pages of the then editions of Odesa are identified. It is shown that the study of publications on the origins of racism and discrimination in a civilized society is one of the most important tasks today. The publications of the South of Ukraine themselves are an important historical source according to the existing ideas of the region's population about the development of the United States. The Odesa newspapers published various news about everyday life of the population abroad. The article indicates that it was difficult for journalists of that time to objectively present information through the political situation. Despite the conditions in which the journalists had to work, they objectively wrote about the struggle of African Americans for rights and freedoms. The use of politically incorrect words and expressions was revealed, but for that time it was the norm. The article explores the materials of the Odesa press of the second half of the 19th century as the main source of foreign news in the South of Ukraine. In the period under review, the importance of periodicals in the social, political and cultural life of Ukraine was extremely important. Any event immediately became known and appeared in the press.

Elsa Speak App: Automatic Speech Recognition (ASR) for Supplementing English Pronunciation Skills

Nowadays, artificial intelligence (AI) became a special concern in language teaching for the reason that it can assist and enhance language learning for all levels of education. Again, it had beneficial roles for supplementing language teaching like ELSA Speak App one of Automatic Speech Recognition (ASR) used for teaching pronunciation. It studied how students heard, voiced, uttered, vocalized, and asserted the English words in the oral language, but the students often pronounced incorrect words with the result that the uttered words had faulty meaning. This study aimed to carry out English Language Speech Assistant (ELSA) Speak App to improve English language pronunciation skills to higher education learners that were the English Department Students of Nahdlatul Ulama University of Yogyakarta (UNU). The data were collected using a test of pronunciation and interview. The researcher also taught in the classroom. The results showed that ELSA Speak can increase the students’ pronunciation skills. It can be seen from the average scores obtained from the teaching cycles from two to four in grade. Clearly, ELSA Speak helped the students pronounce diverse words more easily and comprehensively. Also, the available features offered by this app like instant feedback enabled the students to pronounce precisely. In conclusion, ELSA Speak can improve the students’ pronunciation skills well and effectively. Indeed, it can motivate the students to engage in learning to pronounce.

PROBLEMS AND APPROACHES IN CHILDREN FIRST LANGUAGE ACQUISITION AT AGE 1-3 YEARS OLD IN YEMEN

&lt;p&gt;This study argues the problems and approaches in acquiring the first language at age 1-3 years by children in Yemen. Some stages like pre-production, early production, stage of the first-word stage, beginning fluency in its beginning intermediate, and advanced level are treated in this study. To solve the problem, theories relating to acquiring the first language are employed, these theories include (Varshney, 2003), (Lyons, 1981), (Bolinger, 2002), (Chomsky, 2009), (Gleason J. B., 1998), (Steinberg D. D., 2003), (Fromkin, 1983), (Langacker, 1973), (Wells, Children’s Language and Learning, 1980). This study is carried out by employing descriptive qualitative research. The target children are from Yemen aged 1-3 years old. The researcher’s daughter and son are taken as a subject for this study. Then he observed them for a long time and recorded them by using video aids that help him to collect data. After collecting the data, the researcher notices some problems regarding acquiring the first language. These problems include errors in speech sounds, incorrect words, reproduction, duplication, rectification, specifying the question, naming by experience, single-word response. He also finds out a number of approaches regarding acquiring the first language like pre-production, early production, first word, beginning fluency stage, intermediate fluency stage, and advanced fluency stage. Finally, Children first language acquisition can be arisen by exposing it to their parents and others surrounding them.&lt;strong&gt;&lt;/strong&gt;&lt;/p&gt;

Raising Students’ Vocabulary Mastery Using Selected Game: Using Chalk Down, Turn-around for EYL Students

Based on the 2013 curriculum, students are expected to master common vocabulary in academic text. However, in reality, the above goal has not been achieved. It can be proven by the fact that students still have a common vocabulary and using incorrect words. During the pre-observation, students often experience learning obstacles while learning English. The purpose of this study was to determine the use of Chalk Down, Turn Around Game for improving students’ vocabulary mastery at grade VIII SMP Negeri 1 Bambang. This study used a pre-experimental method in the form of one class given a pre-test and post-test. The number of students involved 20 students and selected using the total sampling technique. The results of the data showed that the mean score on the post-test is higher than the pre-test (85&gt; 60.25), and the t-test value is higher than the t-table score (11,049&gt; 2,092) at the level of increase of 0.05 with degree freedom (df) 19. It means that the use of the Chalk Down, Turn Around Game can improve students' vocabulary.