MUET ESSAY

In today's ever-changing world, the English language has been placed to be one of the important languages in communication, as well as the main medium of instruction in education. To be in line with this acknowledgement, the Ministry of Education of Malaysia has placed the importance of mastering the English language and has highlighted the need for students to excel in the language as early as in the primary school years. To reach this aim, the Ministry of Education has introduced the GPMP (Gred Purata Mata Pelajaran/ Subject Grade Point Average) target for English to schools in every state to ensure every school performs well in the subject. However, there are still a handful of schools that are unable to pass this GPMP target for English due to the poor performance of their students in the subject. Therefore, this study aims to investigate the factors that interfere with the students' performance in learning English. The participants of this study were the 116 students from the seven schools that were unable to reach the state's GPMP in the English language. This study used questionnaire as the instrument to explore the factors that hindered them from performing well in the subject. Overall, the findings show that there are three major themes that contribute to the students' difficulties in the English language learning; the lack of English vocabulary, the influence of the first language and the socio-economic status of the family. The findings and discussion of this study would be able to help the pertinent parties such as teachers, school administrators and parents to make decisions for the learners' improvement in their English language learning.

The presence of several Innovative Pedagogy principles in Common European of References (CEFR) particularly through the inclusion of Differentiation Strategies and Formative Assessment has encouraged a creative and innovative teaching and learning environment among Malaysian English teachers and pupils. There is a lack of priority towards an effective practice of Differentiation Strategies and Formative Assessment in a CEFR lesson due to the focus on merely achieving the learning standards. CEFR has received a mixed review from educators and test developers due to a lack of understanding of the framework. This situation has directly affected the pedagogy process, especially in the insertion of differentiation strategies and formative assessment. In relation to this, this paper aims to present the perspective of implementing innovative teaching and learning approaches in CEFR while highlighting the embodiment of the principles of innovative pedagogy through Differentiation Strategies and Formative Assessment recommended in the Schemes of Work. By understanding this, educators could make attempts in implementing these new approaches in their own classroom context, especially to the current generation of pupils in order to promote more active learners. It is hoped that these new approaches could yield better achievements among the pupils.

Language Learning Strategies (LLSs) are crucial and helpful in the learning of English as a second language (ESL). However, it appeared that the unawareness of their LLSs has affected ESL learners in the area of this study in their learning of English language. Therefore, this study was conducted to identify the LLSs used among ESL learners in a rural primary school. Data were collected using a survey research design through questionnaire. A set of questionnaire comprises four parts were given to a group of 30 Year 6 pupils in a rural area primary school in Borneo, Malaysia. Data collected showed that majority of or almost all respondents used the listening strategies listed in the questionnaire. Meanwhile, only limited speaking strategies being used by respondents in this study. This is because there was a big number of respondents who felt that speaking strategies are not significant to them due to limited usage of English language. These findings illuminated that majority of the respondents did not use some speaking strategies listed in the questionnaire when learning the English language. For translation strategies, data collected showed that the respondents used the translation strategies listed in the questionnaire when learning the English language. The findings of this study benefit the language teachers in primary schools in a way that these findings appear to be a good reference for them to be aware of their teaching approaches and to ensure their approaches suit learners' LLSs. Overall, the findings of this study showed that listening and translation strategy supported learners more in their language learning.

The e-rater ® automated scoring engine used at Educational Testing Service (ETS) scores the writing quality of essays. In the current practice, e-rater scores are generated via a multiple linear regression (MLR) model as a linear combination of various features evaluated for each essay and human scores as the outcome variable. This study evaluates alternative scoring models based on several additional machine learning algorithms, including support vector machines (SVM), random forests (RF), and k-nearest neighbor regression (k-NN). The results suggest that models based on the SVM algorithm outperform MLR models in predicting human scores. Specifically, SVM-based models yielded the highest agreement between human and e-rater scores. Furthermore, compared with MLR, SVM-based models improved the agreement between human and e-rater scores at the ends of the score scale. In addition, the high correlation between SVM-based e-rater scores with external measures such as examinee's scores on the other parts of the test provided some validity evidence for SVM-based e-rater scores. Future research is encouraged to explore the generalizability of these findings.

With the rapid advancement in educational technology, automated essay grading has emerged as a significant area of research. Automated essay grading systems aim to provide a fair, consistent, and efficient alternative to traditional manual grading methods. These systems score student essays by extracting various features, including grammatical and textual information. To address the challenges of high dimensionality, redundancy issues, computational complexity and feature selection, we are improving the Long Short-Term
Memory-based Augmented Language Model (LSTM-ALM) by optimizing it with the Latent Dirichlet Allocation (LDA) algorithm. By incorporating Natural Language Processing (NLP) techniques, our LDA-optimized LSTM-ALM model enhances the efficiency and accuracy of automated essay scoring, offering a robust solution to the limitations of traditional methods.
Keywords: Automated Essay Grading, Educational Technology, Natural Language Processing, Artificial intelligence, LSTM

Automated Essay Scoring (AES) is a service or software that can predictively grade essay based on a pre-trained computational model. It has gained a lot of research interest in educational institutions as it expedites the process and reduces the effort of human raters in grading the essays as close to humans’ decisions. Despite the strong appeal, its implementation varies widely according to researchers’ preferences. This critical review examines various AES development milestones specifically on different methodologies and attributes used in deriving essay scores. To generalize existing AES systems according to their constructs, we attempted to fit all of them into three frameworks which are content similarity, machine learning and hybrid. In addition, we presented and compared various common evaluation metrics in measuring the efficiency of AES and proposed Quadratic Weighted Kappa (QWK) as standard evaluation metric since it corrects the agreement purely by chance when estimate t...

Heterogeneous networks with dense small cell deployment cause high inter-cell interference. To mitigate inter-cell interference, many dynamic interference coordination (IC) approaches have been proposed. For dense deployment with high interference among cells, traditional forward link dynamic IC can cause high air-link packet error rate (PER), due to changing RF conditions and inaccurate channel quality indication (CQI), reducing or even eliminating any improvements achieved by interference coordination. This research describes dynamic centralized interference coordination approaches to improve forward link performance for both infinite buffer (i.e., FTP download) and delay sensitive (i.e., VOIP) traffic for Enterprise LTE femto cell networks. These approaches employ a two-level scheme: at the central level, the algorithm aims to maximize the network utility, by partitioning the network into clusters and exhaustively searching for optimized resource allocation solutions among femtocells (femto access points) within each cluster; at femtocell level, the algorithm employs existing static approaches, such as soft frequency reuse (SFR), partial frequency reuse (PFR) to reduce air-link PER, thus called utilitybased SFR (USFR) and utility-based PFR approach (UPFR). The cell throughput and UE's cell edge throughput (Tput) of infinite buffer traffic, packet loss rate (PLR) of VOIP traffic, have been characterized and compared between proposed approaches and previously proposed methods.

As one of the most widely spoken languages in the world, English in Malaysia is also an important language for both national and global operations widely used in the aspects of business, education, and international relations. Due to such importance, there is constant concern towards the English language performance of students at all levels of education, and it is one of the issues addressed in the Malaysian Education Blueprint 2013-2025. This paper sought to answer the question of 'What are the factors contributing to students' English language performance?', with specific reference to students from the religious education background at an Islamic-based public university in Malaysia. Derived from a larger study, this paper reported the qualitative findings gathered from interview sessions with a total of 47 English teachers, English and content subject lecturers as well as students regarding the factors contributing to the students' English language performance. Findings revealed 4 main themes of internal features or inner components that influence the students' English achievement which are: (1) teacher attributes (2) learner determination (3) learner beliefs / attitudes and (4) 'I don't understand'. The findings of this study reveal the needs to explore new areas to address the issue of students' English language performance, particularly those of religious background education. It also has broader implications for the global community, which makes it critical for educators and policymakers to collaborate in developing new strategies and resources to help learners overcome the challenges and develop the language skills they need to thrive in today's interconnected world. With that, it is hoped that students are able to access the same opportunities for success as their peers, regardless of their linguistic or cultural background.

Automated Essay Scoring (AES) systems involve using a specially designed computing program to mark students' essays. It is a form of online assessment supported by natural language processing (NLP). These systems seek to exploit advanced technologies to reduce the time and effort spent on the exam scoring process. These systems have been applied in several languages, including Arabic. Nevertheless, the applicable NLP techniques in Arabic AES are still limited, and further investigation is needed to make NLP suitable for Arabic to achieve human-like scoring accuracy. Therefore, this comparative empirical experimental study tested two wordembedding deep learning approaches, namely BERT and Word2vec, along with a knowledge-based similarity approach; Arabic WordNet. The study used the Cosine similarity measure to provide optimal student answer scores. Several experiments were conducted for each of the proposed approaches on two available Arabic short answer question datasets to explore the effect of the stemming level. The quantitative results of this study indicated that advanced models of contextual embedding can improve the efficiency of Arabic AES as the meaning of words can differ in the different contexts. Therefore, serve as a catalyst for future research based on contextual embedding models, as the BERT approach achieved the best Pearson Correlation (.84) and RMSE (1.003). However, this research area needs further investigation to increase the accuracy of Arabic AES to become a practical online scoring system.

Previous research findings revealed that students studying Arabic at Umaru Ali Shankafi Polytechnic and Shehu Shagari College of Education Sokoto State, Nigeria, have weak reading ability in the Arabic language. Significantly, which leads us to believe that this issue is real. The study's objectives are to examine the perspectives of Arabic language teachers on the use of Arabic loanwords in the teaching of reading skills and to determine the relationship between the teaching of reading skills with the use of Arabic loanwords and motivating students toward learning Arabic language. The study uses both a descriptive and an inferential quantitative technique. The study sample was made up of (50) teachers of Arabic from the Shehu Shagari College of Education and the Umaru Ali Shinkafi Polytechnic Sokoto. The findings for the first objective established the impact of Arabic loanwords in teaching reading skills to Arabic learners, with the arithmetic mean value ranging between (3.71-4.16). The Spearman correlation coefficient revealed the findings of the second objective which highlighted the positive relationship between the use of Arabic loanwords and encouragement to learn, where the statistical significance came at the level (α = 0.05), which achieves the positive direct relationship from (x) (Arabic loanwords) to (y) encouragement to learning of the Arabic language. In light of the findings, the study recommended employing these loan words in reading skills to Hausa speaker's learners of the Arabic language.

Research suggests "write-to-learn" tasks improve learning outcomes, yet constructed-response methods of formative assessment become unwieldy with large class sizes. This study evaluates natural language processing algorithms to assist this aim. Six short-answer tasks completed by 1,935 students were scored by several human raters using a detailed rubric and an algorithm. Results indicate substantial inter-rater agreement using quadratic weighted kappa for rater pairs (each QWK > 0.74) and group consensus (Fleiss' Kappa = 0.68). Additionally, intra-rater agreement was estimated for one rater who had scored 178 responses seven years prior (QWK = 0.88). With compelling rater agreement, the study then pilots cluster analysis of response text toward enabling instructors to ascribe meaning to clusters as a means for scalable formative assessment.

English as a medium of instruction (EMI) has been increasingly implemented in schools globally, including in non-English-speaking countries. This implementation aims to prepare students for the global workforce and improve their proficiency in English as a lingua franca. However, the effectiveness of EMI has been a widely debated topic in education. This systematic literature review aims to provide insights into the implementation of EMI in schools by analyzing studies published in academic journals. The review shows that EMI implementation varies depending on the context and stakeholders involved. It highlights the importance of considering factors such as teachers' qualifications, students' language proficiency, and curriculum design. The review also reflects the impact of EMI on students' academic achievement, language proficiency, and socio-cultural development. The review suggests that implementing EMI can result in improved language proficiency and academic outcomes when appropriately designed and executed. However, the review also highlights concerns about the potential negative impact of EMI on the cultural and linguistic identity development of non-native English speakers.

English as a medium of instruction (EMI) has been increasingly implemented in schools globally, including in non-English-speaking countries. This implementation aims to prepare students for the global workforce and improve their proficiency in English as a lingua franca. However, the effectiveness of EMI has been a widely debated topic in education. This systematic literature review aims to provide insights into the implementation of EMI in schools by analyzing studies published in academic journals. The review shows that EMI implementation varies depending on the context and stakeholders involved. It highlights the importance of considering factors such as teachers' qualifications, students' language proficiency, and curriculum design. The review also reflects the impact of EMI on students' academic achievement, language proficiency, and socio-cultural development. The review suggests that implementing EMI can result in improved language proficiency and academic outcomes when appropriately designed and executed. However, the review also highlights concerns about the potential negative impact of EMI on the cultural and linguistic identity development of non-native English speakers.

The National High School Exam (ENEM) in Brazil is a test applied annually to assess students before entering higher education. On average, over 7.5 million students participate in this test. In the same sense, large educational groups need to conduct tests for students preparing for ENEM. For correcting each essay, it is necessary at least two evaluators, which makes the process time consuming and very expensive. One alternative for substantially reducing the cost and speed up the correction of essays is to replace one human evaluator by an automated process. This paper presents a computational approach for essays correction able to replace one human evaluator. Techniques based on feature engineering and deep learning were compared, aiming to obtain the best accuracy among them. It was found that is possible to reach accuracy indexes close to 100% in the most frequent classes that comprise near 80% of the essays set.

The National High School Exam (ENEM) in Brazil is a test applied annually to assess students before entering higher education. On average, over 7.5 million students participate in this test. In the same sense, large educational groups need to conduct tests for students preparing for ENEM. For correcting each essay, it is necessary at least two evaluators, which makes the process time consuming and very expensive. One alternative for substantially reducing the cost and speed up the correction of essays is to replace one human evaluator by an automated process. This paper presents a computational approach for essays correction able to replace one human evaluator. Techniques based on feature engineering and deep learning were compared, aiming to obtain the best accuracy among them. It was found that is possible to reach accuracy indexes close to 100% in the most frequent classes that comprise near 80% of the essays set.

In this study, we present the first spell error corpus for the Indonesian Language (SPECIL). This corpus provides a comprehensive resource for researchers and practitioners to detect and correct spelling errors in Bahasa Indonesia (Indonesian). It should be emphasized that currently, there is no recognized corpus for identifying spelling mistakes in the Indonesian language that has been officially released or made accessible. This study also provides a systematic literature review to identify resources and methodologies for building a corpus for spelling error detection and correction in Indonesia. A corpus was created using a combination of manual and automatic methods. The results of this study are a review of publications relating to corpora and spelling, the novel algorithm of six types of spelling errors, and the production of a corpus comprising over 180,000 tokens in 21,500 sentences, including non-word, real-word, and punctuation errors. Using the developed corpus, various Natural Language Processing (NLP) models, including spell checkers and language models, can be trained and tested to enhance their accuracy and effectiveness in identifying and rectifying errors in Indonesian texts.Moreover, the corpus can be used to develop and evaluate new algorithms and techniques for spelling error detection and correction in Indonesia. The SPECIL corpus is publicly available and accessible.It is expected that SPECIL will inspire further research in this area and facilitate the development of more accurate and effective spelling error detection and correction tools in Indonesian language. INDEX TERMS Corpus, Indonesian language, natural language processing, spell.

As one of the most widely spoken languages in the world, English in Malaysia is also an important language for both national and global operations widely used in the aspects of business, education, and international relations. Due to such importance, there is constant concern towards the English language performance of students at all levels of education, and it is one of the issues addressed in the Malaysian Education Blueprint 2013-2025. This paper sought to answer the question of 'What are the factors contributing to students' English language performance?', with specific reference to students from the religious education background at an Islamic-based public university in Malaysia. Derived from a larger study, this paper reported the qualitative findings gathered from interview sessions with a total of 47 English teachers, English and content subject lecturers as well as students regarding the factors contributing to the students' English language performance. Findings revealed 4 main themes of internal features or inner components that influence the students' English achievement which are: (1) teacher attributes (2) learner determination (3) learner beliefs / attitudes and (4) 'I don't understand'. The findings of this study reveal the needs to explore new areas to address the issue of students' English language performance, particularly those of religious background education. It also has broader implications for the global community, which makes it critical for educators and policymakers to collaborate in developing new strategies and resources to help learners overcome the challenges and develop the language skills they need to thrive in today's interconnected world. With that, it is hoped that students are able to access the same opportunities for success as their peers, regardless of their linguistic or cultural background.

The most effective technique for improving students writing skills is for them to get immediate instructor's feedback and as often as possible. This, however, significantly expands the workload of the instructors. There is a growing need for automated systems to help students draft essays. Automated essay evaluation is increasingly popular in the field of educational evaluation technology. In this work, we present an automatic evaluator of student essays in Arabic language, a system that is modeled on the scheme followed by the schoolteachers in Riyadh, the capital of Saudi Arabia. The main criteria for assessing the essays is: language proficiency; the structure of the essay; and the content which should match the topic. With this in mind, we developed a scheme that relies on latent semantic analysis, and rhetorical structure theory. The system was tested on over 300 different essays, all handwritten by schoolchildren covering various subjects. The performance was measured by machine-human correlation in grading. Our system achieved an overall correlation of 0.79 with the teachers' evaluation.

The aim of this study was to analyze students' readiness and interest in blended learning using the latest technology in Principles of Accounting subject in schools of the southern zone of Malaysia. The study used a quantitative approach with the study design on a survey basis based on a questionnaire. It involved 400 form four students from 30 secondary schools in the states of Johor, Melaka and Negeri Sembilan. The selection of schools and respondents was made using simple random sampling. The data was analyzed using descriptive tests. The findings showed that students' readiness for blended learning in Principles of Accounting subject was at a high level (mean = 4.19) while students' interest in learning Accounting Principles through the blended learning method had a high level as well (mean = 4.33). Overall students are ready to accept blended learning methods in the process of learning Principles of Accounting subject. As an implication of the study, teachers of these subjects can take the opportunity to provide a blended learning method using technology that can increase students' interest, understanding and mastery compared to conventional methods that are usually teacher-centered.

Automatic essay grading (AEG) is a process in which machines assign a grade to an essay written in response to a topic, called the prompt. Zero-shot AEG is when we train a system to grade essays written to a new prompt which was not present in our training data. In this paper, we describe a solution to the problem of zero-shot automatic essay grading, using cognitive information, in the form of gaze behaviour. Our experiments show that using gaze behaviour helps in improving the performance of AEG systems, especially when we provide a new essay written in response to a new prompt for scoring, by an average of almost 5 percentage points of QWK.

The advancement in computer and telecommunication technology has been effecting changes in every area, including on education area, where the learning system is shifting from conventional way to a future technology way called as "E-Learning". E-Learning concept means that all learning activities are delivered electronically, also the assessment can be done using online communication, from answering the quizzes, and getting the final score/grading. This thing will give benefit to the instructors and students because to the time consumed for the assessment will be much shorter, more effective and efficient. Another benefit from this system is that the coring will be much faster and accurate. And also this system can handle bigger class with more students. The assessment model whether by conventional or E-Learning, based on two main formats: an Objective Format and Essay Format. Examination in essay is an indicator to determine student's capability where choices of the answers are not provided. The methods used in automated essay grading system nowadays are still under research because they need to follow some specific rules. Some of the methods are Natural Language Processing, Statistical Approach, Bayesian Text Classification, and Latent Semantic Analysis. Essay question is a format where answer is not provided, students have to answers with correct sentences, and this condition could lead to the different answer which comes up from every participant of the test. On the previous system called SIMPLE-O version 1.0, the algorithm and weighted value was already implemented. The process of essay assessment was running well and also the final mark of the assessment was successfully generated by the system. This research will build a system called SIMPLE-O, using LSA method, based on Indonesian Language and add a weighted on words that considered important from chosen keyword. The weight value will be given by multiplying value by 2. Aside, on this system, similar words will be carefully watched, since it becomes a significant factor for the system using Indonesian language. On the previous research, this system generated the result of Agreement value for human raters in small-size class is 68.90%-94.64% and middle-size class is 77.18%-98.42%. This enhancement research had carried out to implement different weighted value by using weight scheme SICBI, where the keyword is adjusted to improve the Agreement value of human raters into 73.12%-94.33% for middle-size class. Besides, the SVD process is also being improved by implementing better and more efficient algorithm, so the time for computation process will be shorter and more efficient, which is 6.47 times faster. In addition, this research also improves the aspect of user comfort and friendliness to interact with the system and also to reduce the number of possibility mistypes of student answers.

As one of the most widely spoken languages in the world, English in Malaysia is also an important language for both national and global operations widely used in the aspects of business, education, and international relations. Due to such importance, there is constant concern towards the English language performance of students at all levels of education, and it is one of the issues addressed in the Malaysian Education Blueprint 2013-2025. This paper sought to answer the question of 'What are the factors contributing to students' English language performance?', with specific reference to students from the religious education background at an Islamic-based public university in Malaysia. Derived from a larger study, this paper reported the qualitative findings gathered from interview sessions with a total of 47 English teachers, English and content subject lecturers as well as students regarding the factors contributing to the students' English language performance. Findings revealed 4 main themes of internal features or inner components that influence the students' English achievement which are: (1) teacher attributes (2) learner determination (3) learner beliefs / attitudes and (4) 'I don't understand'. The findings of this study reveal the needs to explore new areas to address the issue of students' English language performance, particularly those of religious background education. It also has broader implications for the global community, which makes it critical for educators and policymakers to collaborate in developing new strategies and resources to help learners overcome the challenges and develop the language skills they need to thrive in today's interconnected world. With that, it is hoped that students are able to access the same opportunities for success as their peers, regardless of their linguistic or cultural background.

Good writing skills are essential in the education system at all levels. However, the evaluation of essays is labor intensive and can entail a subjective bias. Automated Essay Scoring (AES) is a tool that may be able to save teacher time and provide more objective evaluations. There are several successful AES systems for essays in English that are used in large scale tests. Supervised machine learning algorithms are the core component in developing these systems. In this project four AES systems were developed and evaluated. The AES systems were based on standard supervised machine learning software, i.e., LDAC, SVM with RBF kernel, polynomial kernel and Extremely Randomized Trees. The training data consisted of 1500 high school essays that had been scored by the students' teachers and blind raters. To evaluate the AES systems, the agreement between blind raters' scores and AES scores was compared to agreement between blind raters' and teacher scores. On average, the agreement between blind raters and the AES systems was better than between blind raters and teachers. The AES based on LDAC software had the best agreement with a quadratic weighted kappa value of 0.475. In comparison, the teachers and blind raters had a value of 0.391. However the AES results do not meet the required minimum agreement of a quadratic weighted kappa of 0.7 as defined by the US based nonprofit organization Educational Testing Services. Sammanfattning Jag har utvecklat och utvärderat fyra system för automatisk betygsättning av uppsatser (AES). LDAC, SVM med RBF kernel, SVM med Polynomial kernel och "Extremely Randomized trees" som är standard klassificerarprogramvaror har använts som grunden för att bygga respektivt AES system.

As one of the most widely spoken languages in the world, English in Malaysia is also an important language for both national and global operations widely used in the aspects of business, education, and international relations. Due to such importance, there is constant concern towards the English language performance of students at all levels of education, and it is one of the issues addressed in the Malaysian Education Blueprint 2013-2025. This paper sought to answer the question of 'What are the factors contributing to students' English language performance?', with specific reference to students from the religious education background at an Islamic-based public university in Malaysia. Derived from a larger study, this paper reported the qualitative findings gathered from interview sessions with a total of 47 English teachers, English and content subject lecturers as well as students regarding the factors contributing to the students' English language performance. Findings revealed 4 main themes of internal features or inner components that influence the students' English achievement which are: (1) teacher attributes (2) learner determination (3) learner beliefs / attitudes and (4) 'I don't understand'. The findings of this study reveal the needs to explore new areas to address the issue of students' English language performance, particularly those of religious background education. It also has broader implications for the global community, which makes it critical for educators and policymakers to collaborate in developing new strategies and resources to help learners overcome the challenges and develop the language skills they need to thrive in today's interconnected world. With that, it is hoped that students are able to access the same opportunities for success as their peers, regardless of their linguistic or cultural background.

Globally, English is the second language of many countries, including Malaysia. It is a compulsory subject at all levels of the education system, including preschool, primary, secondary and tertiary education. At tertiary level, Malaysian University English Test (MUET) is a mandatory requirement for admission and graduation for Malaysian public and private universities. However, the record shows that a number of students have not met this requirement and thus are not eligible to graduate. This issue has raised concern among educators of Universiti Sains Islam Malaysia (USIM) to unpack the factors that might have hindered students from successfully meeting the required band as well as factors that might have assisted some students to have achieved the target.
The goal of this project is to propose a framework for improving students’ MUET results. This framework is intended to serve as a guide for relevant parties (university management and language practitioners) to take actions that will assist students in improving their English language proficiency and thus their MUET band, allowing them to achieve the required band for graduation.
The first chapter, The Background, explains the reader about the background of the project. It begins with an overview of English language education practices in Malaysia, including MUET. It continues with a description of the scenario of students’ admission to USIM, followed by the project’s significance particularly in the issues of English language and MUET in USIM. The chapter then proceeds with a description of best practices in other universities.
The second chapter, The Project, depicts the design of the framework. Adopting a mixed-methods design, the authors have used several methods of obtaining information to achieve the project objectives. It is done through document analysis, survey and interviews with various parties. The project’s overall findings are used to create a framework for improving students’ English language performance.
The factors that contributed to USIM students’ MUET results are addressed in the third chapter, The Factors and built based on feedback from various parties involved. Eight features are identified as the contributing factors to students’ MUET results and classified as external and internal factors.
The book continues with Chapter 4, The Framework. From the findings, the authors have developed the framework to enhance students’ English language performance. On top of that, it is also relevant to be used or applied by any institution as a guideline in improving students’ English performance.
Another interesting discussion follows in Chapter 5, The Plan. Based on the framework mentioned in the previous chapter, a plan has been developed namely Strategic Plan of English Language (StraPEL) to enhance students’ English performance. It demonstrates how the five elements of the framework are being implemented as institutional thrusts for English language learning in USIM context.
To conclude, the established framework is significant as a way forward that provides a guideline for the English language and education practitioners in enhancing students’ English language performance.

The most effective technique for improving students writing skills is for them to get immediate instructor's feedback and as often as possible. This, however, significantly expands the workload of the instructors. There is a growing need for automated systems to help students draft essays. Automated essay evaluation is increasingly popular in the field of educational evaluation technology. In this work, we present an automatic evaluator of student essays in Arabic language, a system that is modeled on the scheme followed by the schoolteachers in Riyadh, the capital of Saudi Arabia. The main criteria for assessing the essays is: language proficiency; the structure of the essay; and the content which should match the topic. With this in mind, we developed a scheme that relies on latent semantic analysis, and rhetorical structure theory. The system was tested on over 300 different essays, all handwritten by schoolchildren covering various subjects. The performance was measured by machine-human correlation in grading. Our system achieved an overall correlation of 0.79 with the teachers' evaluation.

We study the effects of data size and quality on the performance on Automated Essay Scoring (AES) engines that are designed in accordance with three different paradigms; A frequency and hand-crafted feature-based model, a recurrent neural network model, and a pretrained transformer-based language model that is fine-tuned for classification. We expect that each type of model benefits from the size and the quality of the training data in very different ways. Standard practices for developing training data for AES engines were established with feature-based methods in mind, however, since neural networks are increasingly being considered in a production setting, this work seeks to inform us as to how to establish better training data for neural networks that will be used in production.

The most productive strategy to improve students' ability to write is to have direct and as much as possible teacher input. However, the workload of the teacher is greatly increased. Automated systems are increasingly required to help students write essays. In the field of educational assessment technology, automated test evaluation is becoming more and more common. We present a framework that is modelled on the programme, following which the school-teachers in the BPDAV School and Govt. High School Hyderabad, Telangana, India present the automatic evaluator of student essays in the Telugu language.  Language skills; the structure of the essay and the contents that fit the subject are the principal requirements for evaluating the essays. In this context, we have established a scheme focused on latent semantical analysis and the theory of rhetorical structure. The method has been evaluated in more than 600 different essays, written in different manuscripts by schoolchildren. Over...

The task of human-based grading essays has been a big trouble due to the long time it takes, subjectiveness in its nature, difficulty it involves, and focus it requires to minimize the possibility of errors as much as possible. As it is the requirement for most academic institutions then an essay must be evaluated at least twice by different examiners. These reasons push us toward investigating and exploring more into this scope. The Computer Science research pipeline (which involves Data collection, preprocessing, defining and training the model, and finally evaluating the proposed model) was followed. The dataset of human graded essays was obtained from Kaggle website. These essays were divided into 8 sets based on context and domain. We first cleaned and pre-processed data using NLP techniques and converted the textual data into numeric using a word-embedding approach. After that, we fed generated vectors as input into the model to predict the scores. The K-Fold approach was employed to tackle the training and testing stages. Finally, the model performance was evaluated using different evaluation metrics. For LSTM, averaged Quadratic Weighted Kappa QWK was used to figure out the correlation between the predicted and original human grade. The evaluation result for the proposed model obtained was 96%. The model was then used to deploy a web application that allows users to submit their essay work and have it graded instantly just like a simple multiple choice online test.

Traditional automated essay scoring systems rely on carefully designed features to evaluate and score essays. The performance of such systems is tightly bound to the quality of the underlying features. However, it is laborious to manually design the most informative features for such a system. In this paper, we develop an approach based on recurrent neural networks to learn the relation between an essay and its assigned score, without any feature engineering. We explore several neural network models for the task of automated essay scoring and perform some analysis to get some insights of the models. The results show that our best system, which is based on long short-term memory networks, outperforms a strong baseline by 5.6% in terms of quadratic weighted Kappa, without requiring any feature engineering.

The advancement in computer and telecommunication technology has been effecting changes in every area, including on education area, where the learning system is shifting from conventional way to a future technology way called as "E-Learning". E-Learning concept means that all learning activities are delivered electronically, also the assessment can be done using online communication, from answering the quizzes, and getting the final score/grading. This thing will give benefit to the instructors and students because to the time consumed for the assessment will be much shorter, more effective and efficient. Another benefit from this system is that the coring will be much faster and accurate. And also this system can handle bigger class with more students. The assessment model whether by conventional or E-Learning, based on two main formats: an Objective Format and Essay Format. Examination in essay is an indicator to determine student's capability where choices of the answers are not provided. The methods used in automated essay grading system nowadays are still under research because they need to follow some specific rules. Some of the methods are Natural Language Processing, Statistical Approach, Bayesian Text Classification, and Latent Semantic Analysis. Essay question is a format where answer is not provided, students have to answers with correct sentences, and this condition could lead to the different answer which comes up from every participant of the test. On the previous system called SIMPLE-O version 1.0, the algorithm and weighted value was already implemented. The process of essay assessment was running well and also the final mark of the assessment was successfully generated by the system. This research will build a system called SIMPLE-O, using LSA method, based on Indonesian Language and add a weighted on words that considered important from chosen keyword. The weight value will be given by multiplying value by 2. Aside, on this system, similar words will be carefully watched, since it becomes a significant factor for the system using Indonesian language. On the previous research, this system generated the result of Agreement value for human raters in small-size class is 68.90%-94.64% and middle-size class is 77.18%-98.42%. This enhancement research had carried out to implement different weighted value by using weight scheme SICBI, where the keyword is adjusted to improve the Agreement value of human raters into 73.12%-94.33% for middle-size class. Besides, the SVD process is also being improved by implementing better and more efficient algorithm, so the time for computation process will be shorter and more efficient, which is 6.47 times faster. In addition, this research also improves the aspect of user comfort and friendliness to interact with the system and also to reduce the number of possibility mistypes of student answers.

An automated essay evaluation system is a machine-based approach leveraging long short-term memory (LSTM) model to award grades to essays written in English language. natural language processing (NLP) is used to extract feature representations from the essays. The LSTM network learns from the extracted features and generates parameters for testing and validation. The main objectives of the research include proposing and training an LSTM model using a dataset of manually graded essays with scores. Sentiment analysis is performed to determine the sentiment of the essay as either positive, negative or neutral. The twitter sample dataset is used to build sentiment classifier that analyzes the sentiment based on the student's approach towards a topic. Additionally, each essay is subjected to detection of syntactical errors as well as plagiarism check to detect the novelty of the essay. The overall grade is calculated based on the quality of the essay, the number of syntactic errors, the percentage of plagiarism found and sentiment of the essay. The corrected essay is provided as feedback to the students. This essay grading model has gained an average quadratic weighted kappa (QWK) score of 0.911 with 99.4% accuracy for the sentiment analysis classifier.

Traditionally, computer programming assignments are graded manually by educators. As this task is tedious, timeconsuming and prone to bias, the need for automated grading tool is necessary to reduce the educators' burden and avoid inconsistency and favoritism. Recent researches have claimed that Latent Semantic Analysis (LSA) has the ability to represent human cognitive knowledge to assess essays, retrieving information, classification of documents and indexing. In this paper, we adapt LSA technique to grade computer programming assignments and observe how far it can be applied as an alternative approach to traditional grading methods by human. The grades of the assignments are generated from the cosine similarity that shows how close students' assignments to the model answers in the latent semantic vector space. The results show that LSA is not able to detect orders of computer programming and symbols; however, LSA is able to grade assignments faster and consistently, which avoid bias and reduces the time spent by human.

The manual process of scoring short answers of Arabic essay questions is exhaustive, susceptible to error and consumes instructor's time and resources. This paper explores longest common subsequence (LCS) algorithm as a string-based text similarity measure for effectively scoring short answers of Arabic essay questions. To achieve this effectiveness, the longest common subsequence is modified by developing weight-based measurement techniques and implemented along with using Arabic WordNet for scoring Arabic short answers. The experiments conducted on a dataset of 330 students' answers reported Root Mean Square Error (RMSE) value of 0.81 and Pearson correlation r value of 0.94. Findings based on experiments have shown improvements in the accuracy of performance of the proposed approach compared to similar studies. Moreover, the statistical analysis has shown that the proposed method scores students' answers similar to that of human estimator.

Neural approaches to automated essay scoring have recently shown state-of-theart performance. The automated essay scoring task typically involves a broad notion of writing quality that encompasses content, grammar, organization, and conventions. This differs from the short answer content scoring task, which focuses on content accuracy. The inputs to neural essay scoring models-ngrams and embeddings-are arguably well-suited to evaluate content in short answer scoring tasks. We investigate how several basic neural approaches similar to those used for automated essay scoring perform on short answer scoring. We show that neural architectures can outperform a strong nonneural baseline, but performance and optimal parameter settings vary across the more diverse types of prompts typical of short answer scoring.

We study the effects of data size and quality on the performance on Automated Essay Scoring (AES) engines that are designed in accordance with three different paradigms; A frequency and hand-crafted feature-based model, a recurrent neural network model, and a pretrained transformer-based language model that is fine-tuned for classification. We expect that each type of model benefits from the size and the quality of the training data in very different ways. Standard practices for developing training data for AES engines were established with feature-based methods in mind, however, since neural networks are increasingly being considered in a production setting, this work seeks to inform us as to how to establish better training data for neural networks that will be used in production.

The paper presents a methodology and an application framework (PUAnalyzeThis) that makes use of MeaningCloud API to automatically extract entities, concepts, relations, etc. and calculate scores and grades based on their relevance to a preliminary created topics graph. These topic graphs are either to be created by the teacher or automatically generated by scanning a certain amount of sample texts in the subject area. A prototype has been developed and tested with essays in the field of Computational Linguistics for informatics students at the University of Plovdiv “Paisii Hilendarski”.

The paper presents a methodology and an application framework (PUAnalyzeThis) that makes use of MeaningCloud API to automatically extract entities, concepts, relations, etc. and calculate scores and grades based on their relevance to a preliminary created topics graph. These topic graphs are either to be created by the teacher or automatically generated by scanning a certain amount of sample texts in the subject area. A prototype has been developed and tested with essays in the field of Computational Linguistics for informatics students at the University of Plovdiv “Paisii Hilendarski”.

Deep-learning based Automatic Essay Scoring (AES) systems are being actively used by states and language testing agencies alike to evaluate millions of candidates for life-changing decisions ranging from college applications to visa approvals. However, little research has been put to understand and interpret the black-box nature of deep-learning based scoring algorithms. Previous studies indicate that scoring models can be easily fooled. In this paper, we explore the reason behind their surprising adversarial brittleness. We utilize recent advances in interpretability to find the extent to which features such as coherence, content, vocabulary, and relevance are important for automated scoring mechanisms. We use this to investigate the oversensitivity (i.e., large change in output score with a little change in input essay content) and overstability (i.e., little change in output scores with large changes in input essay content) of AES. Our results indicate that autoscoring models, de...

Automated Essay Scoring (AES) is the use of specialized computer programs to assign grades to essays written in an educational assessment context. It is developed to overcome time, cost, and reliability issues in writing assessment. Most of the contemporary AES are "western" proprietary product, designed for native English speakers, where the source code is not made available to public and the assessment criteria may tend to be associated with the scoring rubrics of a particular English test context. Therefore, such AES may not be appropriate to be directly adopted in Malaysia context. There is no actual software development work found in building an AES for Malaysian English test environment. As such, this work is carried out as the study for formulating the requirement of a local AES, targeted for Malaysia's essay assessment environment. In our work, we assessed a well-known AES called LightSide for determining its suitability in our local context. We use various Machine Learning technique provided by LightSide to predict the score of Malaysian University English Test (MUET) essays; and compare its performance, i.e. the percentage of exact agreement of LightSide with the human score of the essays. Besides, we review and discuss the theoretical aspect of the AES, i.e. its state-of-the-art, reliability and validity requirement. The finding in this paper will be used as the basis of our future work in developing a local AES, namely Intelligent Essay Grader (IEG), for Malaysian English test environment.

We demonstrate that current state-of-theart approaches to Automated Essay Scoring (AES) are not well-suited to capturing adversarially crafted input of grammatical but incoherent sequences of sentences. We develop a neural model of local coherence that can effectively learn connectedness features between sentences, and propose a framework for integrating and jointly training the local coherence model with a state-of-the-art AES model. We evaluate our approach against a number of baselines and experimentally demonstrate its effectiveness on both the AES task and the task of flagging adversarial input, further contributing to the development of an approach that strengthens the validity of neural essay scoring models.

Automated Essay Scoring (AES) is a service or software that can predictively grade essay based on a pre-trained computational model. It has gained a lot of research interest in educational institutions as it expedites the process and reduces the effort of human raters in grading the essays as close to humans’ decisions. Despite the strong appeal, its implementation varies widely according to researchers’ preferences. This critical review examines various AES development milestones specifically on different methodologies and attributes used in deriving essay scores. To generalize existing AES systems according to their constructs, we attempted to fit all of them into three frameworks which are content similarity, machine learning and hybrid. In addition, we presented and compared various common evaluation metrics in measuring the efficiency of AES and proposed Quadratic Weighted Kappa (QWK) as standard evaluation metric since it corrects the agreement purely by chance when estimate t...

The National High School Exam (ENEM) in Brazil is a test applied annually to assess students before entering higher education. On average, over 7.5 million students participate in this test. In the same sense, large educational groups need to conduct tests for students preparing for ENEM. For correcting each essay, it is necessary at least two evaluators, which makes the process time consuming and very expensive. One alternative for substantially reducing the cost and speed up the correction of essays is to replace one human evaluator by an automated process. This paper presents a computational approach for essays correction able to replace one human evaluator. Techniques based on feature engineering and deep learning were compared, aiming to obtain the best accuracy among them. It was found that is possible to reach accuracy indexes close to 100% in the most frequent classes that comprise near 80% of the essays set.

Using simulation tools save the construction cost of the wireless system. It is not necessary to establish the system and test whether it is working properly or not. Therefore, simulation tools are required to be accurate, simple and with the minimum time. From the accuracy perspective, the stacked femtocells built in most available simulators in the literature are widely deployed horizontally. However, this is not the case in the indoor environment where femtocells have to be arranged vertically; such as, in residential towers where apartments are on the top of each other. In this paper, the state of the art of link and system level simulators is introduced. In addition, a three-dimensional (3D) system level simulator is developed in order to help the researcher in the Long Term Evolution (LTE) femtocell field to analyze and investigate more real scenarios of femtocell deployment. The developed simulator allows the researcher to locate a multi-story building in the region of intere...

Automatic Essay Scoring (AES) is defined as the computer technology that evaluates and scores the written essays, aiming to provide computational models to grade essays either automatically or with minimal human involvement. While there are several AES studies in a variety of languages, few of them are focused on the Portuguese language. The main reason is the lack of a corpus with manually graded essays. In order to bridge this gap, we create a large corpus with several essays written by Brazilian high school students on an online platform. All of the essays are argumentative and were scored across five competencies by experts. Moreover, we conducted an experiment on the created corpus and showed challenges posed by the Portuguese language. Our corpus is publicly available at https://github.com/rafaelanchieta/essay.

Essays are one of the most important method for assessing learning and intelligence of a student. Manual essay grading is a time consuming process for the evaluator, a solution to such problem is to make evaluation through computers. Many systems were proposed over past few decades. Each system works on different approach having focus on different attributes. Aim of this paper is to understand and analyze current essay grading systems and compare them primarily focusing on technique used, performance and focused attributes.

Deep-learning based Automatic Essay Scoring (AES) systems are being actively used by states and language testing agencies alike to evaluate millions of candidates for life-changing decisions ranging from college applications to visa approvals. However, little research has been put to understand and interpret the black-box nature of deep-learning based scoring algorithms. Previous studies indicate that scoring models can be easily fooled. In this paper, we explore the reason behind their surprising adversarial brittleness. We utilize recent advances in interpretability to find the extent to which features such as coherence, content, vocabulary, and relevance are important for automated scoring mechanisms. We use this to investigate the oversensitivity (i.e., large change in output score with a little change in input essay content) and overstability (i.e., little change in output scores with large changes in input essay content) of AES. Our results indicate that autoscoring models, de...

This paper focuses on the relationship between different aspects of the linguistic structure of a given language and the complexity of the computer program, whether existing or prospective, that is to be used for the scoring of essays in that language. The first part of the paper discusses common scales used to assess writing products, then briefly describes various methods of Automated Essay Scoring (AES) and reviews several AES programs currently in use. It also presents empirical results attesting to the reliability and validity of these programs, principally with regard to essays written in English. The second part of the paper presents various linguistic features that may vary extensively across languages and examines the ramifications of these features on the complexity of the AES operational system. This analysis is presented chiefly with regard to Hebrew and English, which are used to illustrate the differences that may exist between languages. (Contains 5 tables and 30 refe...