Search results for: learning Maltese as a second language

Authors: Vishal Raj, Noorhan Abbas

Abstract:

Ambiguity in NLP (Natural language processing) refers to the ability of a word, phrase, sentence, or text to have multiple meanings. This results in various kinds of ambiguities such as lexical, syntactic, semantic, anaphoric and referential am-biguities. This study is focused mainly on solving the issue of Lexical ambiguity. Word Sense Disambiguation (WSD) is an NLP technique that aims to resolve lexical ambiguity by determining the correct meaning of a word within a given context. Most WSD solutions rely on words for training and testing, but we have used lemma and Part of Speech (POS) tokens of words for training and testing. Lemma adds generality and POS adds properties of word into token. We have designed a novel method to create an affinity matrix to calculate the affinity be-tween any pair of lemma_POS (a token where lemma and POS of word are joined by underscore) of given training set. Additionally, we have devised an al-gorithm to create the sense clusters of tokens using affinity matrix under hierar-chy of POS of lemma. Furthermore, three different mechanisms to predict the sense of target word using the affinity/similarity value are devised. Each contex-tual token contributes to the sense of target word with some value and whichever sense gets higher value becomes the sense of target word. So, contextual tokens play a key role in creating sense clusters and predicting the sense of target word, hence, the model is named Contextual SenSe Model (CSM). CSM exhibits a noteworthy simplicity and explication lucidity in contrast to contemporary deep learning models characterized by intricacy, time-intensive processes, and chal-lenging explication. CSM is trained on SemCor training data and evaluated on SemEval test dataset. The results indicate that despite the naivety of the method, it achieves promising results when compared to the Most Frequent Sense (MFS) model.

Keywords: word sense disambiguation (wsd), contextual sense model (csm), most frequent sense (mfs), part of speech (pos), natural language processing (nlp), oov (out of vocabulary), lemma_pos (a token where lemma and pos of word are joined by underscore), information retrieval (ir), machine translation (mt)

Procedia PDF Downloads 111

Authors: Carla Michelle Gastelum Knight

Abstract:

In an era where teacher educators and student teachers have almost unrestricted access to all kinds of sources through the internet, a research project carried out with a group of student-teachers has revealed how self-made videos are an exciting new way to motivate and engage students. The project was carried out at Universidad de Sonora, a public university in Northern Mexico, where 39 students of the Bachelor in Arts in English Language Teaching (B.A. in ELT) programme participated creating their own videos. In the process, they worked collaboratively, they exploited their creativity, they were highly motivated and showed more interest in the subject. The videos were shared in a private YouTube channel where students had the opportunity to review their peers’ work and where videos are available at any time for later viewing. This experience has led course instructor to face the challenge of planning and designing meaningful tasks that can and to find ways of exploiting the use of these resources for learning and training purposes.

Keywords: self-made materials, student-teachers, teacher education programme, teacher training

Procedia PDF Downloads 235

Authors: Deokho Kim, Wonwoo Lee, Jaewoong Lee, Teresa Ng, Gun-Ill Lee, Jiwon Jeong

Abstract:

This paper presents a method for improving object search accuracy using a deep learning model. A major limitation to provide accurate similarity with deep learning is the requirement of huge amount of data for training pairwise similarity scores (metrics), which is impractical to collect. Thus, similarity scores are usually trained with a relatively small dataset, which comes from a different domain, causing limited accuracy on measuring similarity. For this reason, this paper proposes a deep learning model that can be trained with a significantly small amount of data, a clustered data which of each cluster contains a set of visually similar images. In order to measure similarity distance with the proposed method, visual features of two images are extracted from intermediate layers of a convolutional neural network with various pooling methods, and the network is trained with pairwise similarity scores which is defined zero for images in identical cluster. The proposed method outperforms the state-of-the-art object similarity scoring techniques on evaluation for finding exact items. The proposed method achieves 86.5% of accuracy compared to the accuracy of the state-of-the-art technique, which is 59.9%. That is, an exact item can be found among four retrieved images with an accuracy of 86.5%, and the rest can possibly be similar products more than the accuracy. Therefore, the proposed method can greatly reduce the amount of training data with an order of magnitude as well as providing a reliable similarity metric.

Keywords: visual search, deep learning, convolutional neural network, machine learning

Procedia PDF Downloads 217

Authors: M. Jebali, M. Jemni

Abstract:

This work is the part of vision based hand gesture recognition system for Natural Human Computer Interface. Hand tracking and segmentation are the primary steps for any hand gesture recognition system. The aim of this paper is to develop robust and efficient hand segmentation algorithm such as an input to another system which attempt to bring the HCI performance nearby the human-human interaction, by modeling an intelligent sign language recognition system based on prediction in the context of dialogue between the system (avatar) and the interlocutor. For the purpose of hand segmentation, an overcoming occlusion approach has been proposed for superior results for detection of hand from an image.

Keywords: HCI, sign language recognition, object tracking, hand segmentation

Procedia PDF Downloads 414

Authors: Ali Reza Sattarzadeh, Ronny J. Kutadinata, Pubudu N. Pathirana, Van Thanh Huynh

Abstract:

Traffic congestion has become a severe worldwide problem, affecting everyday life, fuel consumption, time, and air pollution. The primary causes of these issues are inadequate transportation infrastructure, poor traffic signal management, and rising population. Traffic flow forecasting is one of the essential and effective methods in urban congestion and traffic management, which has attracted the attention of researchers. With the development of technology, undeniable progress has been achieved in existing methods. However, there is a possibility of improvement in the extraction of temporal and spatial features to determine the importance of traffic flow sequences and extraction features. In the proposed model, we implement the convolutional neural network (CNN) and long short-term memory (LSTM) deep learning models for mining nonlinear correlations and their effectiveness in increasing the accuracy of traffic flow prediction in the real dataset. According to the experiments, the results indicate that implementing Conv-LSTM networks increases the productivity and accuracy of deep learning models for traffic flow prediction.

Keywords: deep learning algorithms, intelligent transportation systems, spatiotemporal features, traffic flow prediction

Procedia PDF Downloads 174

Authors: Abdelmenen Abobghala, Mahmud Ahmed, Mohamed Alwaheshi, Anwar Fanan, Meftah Mehdawi, Ahmed Abuhatira

Abstract:

This research aims to predict academic performance and identify weak points in students to aid teachers in understanding their learning needs. Both quantitative and qualitative methods are used to identify difficult test items and the factors causing difficulties. The study uses interventions like focus group discussions, interviews, and action plans developed by the students themselves. The research questions explore the predictability of final grades based on mock exams and assignments, the student's response to action plans, and the impact on learning performance. Ethical considerations are followed, respecting student privacy and maintaining anonymity. The research aims to enhance student engagement, motivation, and responsibility for learning.

Keywords: prediction, academic performance, weak points, understanding, learning, quantitative methods, qualitative methods, formative assessments, feedback, emotional responses, intervention, focus group discussion, interview, action plan, student engagement, motivation, responsibility, ethical considerations

Procedia PDF Downloads 68

Authors: Debneil Saha Roy

Abstract:

Weather forecasting is a complex topic and is well suited for analysis by deep learning approaches. With the wide availability of weather observation data nowadays, these approaches can be utilized to identify immediate comparisons between historical weather forecasts and current observations. This work explores the application of deep learning techniques to weather forecasting in order to accurately predict the weather over a given forecast hori­zon. Three deep neural networks are used in this study, namely, Multi-Layer Perceptron (MLP), Long Short Tunn Memory Network (LSTM) and a combination of Convolutional Neural Network (CNN) and LSTM. The predictive performance of these models is compared using two evaluation metrics. The results show that forecasting accuracy increases with an increase in the complexity of deep neural networks.

Keywords: convolutional neural network, deep learning, long short term memory, multi-layer perceptron

Procedia PDF Downloads 178

Authors: Nora El-Bilawia

Abstract:

This paper discovers the challenges and/or opportunities to implementing multiple intelligence (MI) practices in English as foreign language (EFL) classrooms at Egyptian public schools as part of the government’s educational reform plan. It is found that Egyptian EFL teachers value the use of MI’s ways of teaching as means for active and higher order thinking. However, teachers believed they were underprivileged, as the government did not provide appropriate trainings, tools, or means to integrate MI in their daily lessons. They also conferred challenges they face due to some Egyptian schooling cultural practices. At the end of this chapter, a proposed need for a paradigm shift in the schooling culture in Egypt to implement practical changes in schools to promote hope in education such as the use of MI teaching tools. This study promotes cross-cultural understanding of educational opportunities and efforts for equal learning outcomes around the globe.

Keywords: professional development, schooling culture, acculturation, equitable education

Procedia PDF Downloads 103

Authors: Sheena Christabel Pravin, M. Palanivelan

Abstract:

Speech disfluencies are common in spontaneous speech. The primary purpose of this study was to distinguish linguistic disfluencies from stuttering disfluencies in bilingual Tamil–English (TE) speaking children. The secondary purpose was to determine whether their disfluencies are mediated by native language dominance and/or on an early onset of developmental stuttering at childhood. A detailed study was carried out to identify the prosodic and acoustic features that uniquely represent the disfluent regions of speech. This paper focuses on statistical modeling of repetitions, prolongations, pauses and interjections in the speech corpus encompassing bilingual spontaneous utterances from school going children – English and Tamil. Two classifiers including Hidden Markov Models (HMM) and the Multilayer Perceptron (MLP), which is a class of feed-forward artificial neural network, were compared in the classification of disfluencies. The results of the classifiers document the patterns of disfluency in spontaneous speech samples of school-aged children to distinguish between Children Who Stutter (CWS) and Children with Language Impairment CLI). The ability of the models in classifying the disfluencies was measured in terms of F-measure, Recall, and Precision.

Keywords: bi-lingual, children who stutter, children with language impairment, hidden markov models, multi-layer perceptron, linguistic disfluencies, stuttering disfluencies

Procedia PDF Downloads 218

Authors: Ratima Damkham, Natasha Dejdumrong, Priyakorn Pusawiro

Abstract:

This paper explored the solution in game design to help game designers in the educational game designing using digital educational game model canvas (DEGMC) and digital educational game form (DEGF) based on Outcome-based Education program. DEGMC and DEGF can help designers develop an overview of the game while designing and planning their own game. The way to clearly assess players’ ability from learning outcomes and support their game learning design is by using the tools. Designers can balance educational content and entertainment in designing a game by using the strategies of the Business Model Canvas and design the gameplay and players’ ability assessment from learning outcomes they need by referring to the Constructive Alignment. Furthermore, they can use their design plan in this research to write their Game Design Document (GDD). The success of the research was evaluated by four experts’ perspectives in the education and computer field. From the experiments, the canvas and form helped the game designers model their game according to the learning outcomes and analysis of their own game elements. This method can be a path to research an educational game design in the future.

Keywords: constructive alignment, constructivist theory, educational game, outcome-based education

Procedia PDF Downloads 355

Authors: Emmanuel Barrantes Chaves, Ernesto Rivera Alvarado

Abstract:

This paper presents a method to enable agents for the Terraria game to evaluate algorithms commonly used in general video game artificial intelligence competitions. The usage of the ‘You Only Look Once’ model in the first layer of the process obtains information from the screen, translating this information into a video game description language known as “Video Game Description Language”; the agents take that as input to make decisions. For this, the state-of-the-art algorithms were tested and compared; Monte Carlo Tree Search and Rolling Horizon Evolutionary; in this case, Rolling Horizon Evolutionary shows a better performance. This approach’s main advantage is that a VGDL beforehand is unnecessary. It will be built on the fly and opens the road for using more games as a framework for AI.

Keywords: AI, MCTS, RHEA, Terraria, VGDL, YOLOv5

Procedia PDF Downloads 97

Authors: Reggie Chandra

Abstract:

Trafficaccidents and traffic signal optimization are correlated. However, 70-90% of the traffic signals across the USA are not synchronized. The reason behind that is insufficient resources to create and implement timing plans. In this work, we will discuss the use of a breakthrough Artificial Intelligence (AI) technology to optimize traffic flow and collect 24/7/365 accurate traffic data using a vehicle detection system. We will discuss what are recent advances in Artificial Intelligence technology, how does AI work in vehicles, pedestrians, and bike data collection, creating timing plans, and what is the best workflow for that. Apart from that, this paper will showcase how Artificial Intelligence makes signal timing affordable. We will introduce a technology that uses Convolutional Neural Networks (CNN) and deep learning algorithms to detect, collect data, develop timing plans and deploy them in the field. Convolutional Neural Networks are a class of deep learning networks inspired by the biological processes in the visual cortex. A neural net is modeled after the human brain. It consists of millions of densely connected processing nodes. It is a form of machine learning where the neural net learns to recognize vehicles through training - which is called Deep Learning. The well-trained algorithm overcomes most of the issues faced by other detection methods and provides nearly 100% traffic data accuracy. Through this continuous learning-based method, we can constantly update traffic patterns, generate an unlimited number of timing plans and thus improve vehicle flow. Convolutional Neural Networks not only outperform other detection algorithms but also, in cases such as classifying objects into fine-grained categories, outperform humans. Safety is of primary importance to traffic professionals, but they don't have the studies or data to support their decisions. Currently, one-third of transportation agencies do not collect pedestrian and bike data. We will discuss how the use of Artificial Intelligence for data collection can help reduce pedestrian fatalities and enhance the safety of all vulnerable road users. Moreover, it provides traffic engineers with tools that allow them to unleash their potential, instead of dealing with constant complaints, a snapshot of limited handpicked data, dealing with multiple systems requiring additional work for adaptation. The methodologies used and proposed in the research contain a camera model identification method based on deep Convolutional Neural Networks. The proposed application was evaluated on our data sets acquired through a variety of daily real-world road conditions and compared with the performance of the commonly used methods requiring data collection by counting, evaluating, and adapting it, and running it through well-established algorithms, and then deploying it to the field. This work explores themes such as how technologies powered by Artificial Intelligence can benefit your community and how to translate the complex and often overwhelming benefits into a language accessible to elected officials, community leaders, and the public. Exploring such topics empowers citizens with insider knowledge about the potential of better traffic technology to save lives and improve communities. The synergies that Artificial Intelligence brings to traffic signal control and data collection are unsurpassed.

Keywords: artificial intelligence, convolutional neural networks, data collection, signal control, traffic signal

Procedia PDF Downloads 172

Authors: Hazem Hasan Hushayish

Abstract:

Motivation and teaching strategies are commonly believed to improve students’ communicative competence in English as a foreign language; still, there is not much empirical evidence to support this claim. The present study is intended to focus on the effects of motivational factors and teaching strategies on the communicative competence among the Palestinian undergraduates. In the first phase, one hundred and eighty participants, who are studying English language in three Palestinian universities, answered a questionnaire. The questionnaire included items derived from Gardner’s 2001, 2004, 2006, 2007 Attitude/Motivation Test Battery AMTB and items from Dörnyei 2007 and Guilloteaux and Dörnyei 2008 teaching strategies framework for foreign language classrooms. In the second phase, 6 participants, from the same universities, were interviewed. The quantitative results indicated that participants’ communicative competence is significantly affected by motivation and teaching strategies. Also, the qualitative results indicated that teaching strategies do not directly affect students’ communicative competence, but rather affect their motivation. Consequently, the current study will add substantively to the literature concerning the effects of motivation and teaching strategies in communicative competencies among EFL learners in the Palestinian context, and some suggested procedures and suggestions that help improve learners’ communicative competences.

Keywords: communicative competence, motivation, teaching strategies, Palestinian undergraduates

Procedia PDF Downloads 194

Authors: Tanvir Ahmed, Nahian Fyrose Fahim, Subrata Majumder, Sarker Kibria

Abstract:

English is unanimously recognized as the standard second language in the world, and no one can deny this fact. Many people believe that possessing English proficiency skills is the key to communicating effectively globally, especially for developing countries, which can bring further success to itself on many fronts, as well as to other countries, by ensuring its people worldwide access to education, business, and technology. Bangladesh is a developing country of about 160 million people. A notable number of students in Bangladesh are currently pursuing higher education, especially at the tertiary or collegiate level, in more than 150 public and private universities. English is the dominant linguistic medium through which college instruction and lectures are given to students in Bangladesh. However, many of our students who have only completed their primary and secondary levels of education in the Bangla medium or language are generally in an awkward position to suddenly take and complete many unfamiliar requirements by the time they enter the university as freshmen. As students, they struggle to complete at least 18 courses to acquire proficiency in English. After obtaining a tertiary education certificate, the students could then have the opportunity to acquire a sustainable position in the job market industry; however, many of them do fail, unfortunately, because of poor English proficiency skills. Our study focuses on students in both public and private universities (N=150) as well as education experts (N=30) in Bangladesh. We had prepared two sets of questionnaires that were based upon a literature review on this subject, as we had also collected data and identified the reasons, and arrived at probable solutions to overcoming these problems. After statistical analysis, the study suggested certain remedial measures that could be taken in order to increase student's proficiency in English as well as to ensure their employability potential.

Keywords: tertiary education, English language proficiency, employability, unemployment problems

Procedia PDF Downloads 106

Authors: Mueller Jann, Hoffmann Christian Hugo

Abstract:

Conventional quantitative risk management in banking is a risk factor of its own, because it rests on assumptions such as independence and availability of data which do not hold when rare events of extreme consequences are involved. There is a growing recognition of the need for alternative risk measures that do not make these assumptions. We propose a novel method for modeling the risk associated with investment products, in particular derivatives, by using a formal language for specifying financial contracts. Expressions in this language are interpreted in the category of values annotated with (a formal representation of) uncertainty. The choice of uncertainty formalism thus becomes a parameter of the model, so it can be adapted to the particular application and it is not constrained to classical probabilities. We demonstrate our approach using a simple logic-based uncertainty model and a case study in which we assess the risk of counter party default in a portfolio of collateralized loans.

Keywords: risk model, uncertainty monad, derivatives, contract algebra

Procedia PDF Downloads 578

Authors: Jessica Peart, Sarim Al Zubaidy

Abstract:

This paper examines the main dialogic and reformative aspects that have constituted the developing implementation of an English language module in a common pre-sessional program in Oman, the General Foundation Program (GFP), at the new Military Technological College (MTC), in Oman’s capital, muscat. The MTC is the first of its kind in the country to merge military with academic training and has been running programs since September 2013 over five trimesters to date, receiving external validation and accreditation from the University of Portsmouth (UoP), UK. From this starting point, We will provide context on the parameters that necessitated delivery of this common but specially tailored pre-sessional program at the MTC and outline in detail how the English module with integrated key study skills and personal tutoring support was initially conceived before operations commenced and cooperation between all stakeholders took practical shape. This enquiry traces how stakeholders from students to faculty, college boards and collaborative university partners have considered and redefined the in part static and dynamic boundaries of their larger and smaller scale stakes. With regard to the widely held recognition that pre-sessional students require training in transferable study skills in order to succeed at university, we will chart the subsequent and ongoing adjustments made to the generic, pastoral and integrated elements of that program. Driving this concerted effort has been at base the need for a GFP concerned with three criteria for incoming MTC students cadets, namely to develop candidate’s rounded capacity for intellectual, technical and physical skill as both students and cadets, to generate linguistic proficiency and discerning use of appropriate language registers and to allow personal and collective time for adjustment to a multilayered, brand new environment, while also working within a regulated timeline for academic progression to the MTC diploma or degree levels. The English Department teaching staff’s facilitation of the initial program’s methodologies and timeframe for the GFP English module has garnered a keen and diverse sense of the holistic student cadet experience, which a range of alterations to the program demonstrate. These include alterations to the class types and overall program duration as well as greater multiplicity of exposure within learning environments. In surveying the impact of these composite maneuvers and challenges within a proactive and evolving context of teaching and learning, it is finally demonstrated how student cadet levels of productivity and self-reliance on the one hand and retention issues on the other are being gainfully steered towards progression within a framework for inclusive reciprocal dialogue, gathering thereby civilian and military backgrounds toward uniquely united ends.

Keywords: English module transferable skills, faculty dialogue, governance structure, overarching regulatory agencies

Procedia PDF Downloads 277

Authors: Sara Corvino

Abstract:

The aim of this study is to explore the potential of inclusive learning and assessment strategies to foster students' engagement with historical debates surrounding the field of graphic design. The goal is to respond to the diversity of L4 Graphic Design students, at Nottingham Trent University, in a way that instead of 'lowering standards' can benefit everyone. This research tests, measures, and evaluates the impact of a specific intervention, an assessment task, to develop students' critical visual analysis skills and stimulate a deeper engagement with the subject matter. Within the action research approach, this work has followed a case study research method to understand students' views and perceptions of a specific project. The primary methods of data collection have been: anonymous electronic questionnaire and a paper-based anonymous critical incident questionnaire. NTU College of Business Law and Social Sciences Research Ethics Committee granted the Ethical approval for this research in November 2019. Other methods used to evaluate the impact of this assessment task have been Evasys's report and students' performance. In line with the constructivist paradigm, this study embraces an interpretative and contextualized analysis of the collected data within the triangulation analytical framework. The evaluation of both qualitative and quantitative data demonstrates that active learning strategies and the disruption of thinking patterns can foster greater students' engagement and can lead to meaningful learning.

Keywords: active learning, assessment for learning, graphic design, higher education, student engagement

Procedia PDF Downloads 181

Authors: P. Halder, A. Zaman

Abstract:

It is estimated that heart disease accounts for one in ten deaths worldwide. United States deaths due to heart disease are among the leading causes of death according to the World Health Organization. Cardiovascular diseases (CVDs) account for one in four U.S. deaths, according to the Centers for Disease Control and Prevention (CDC). According to statistics, women are more likely than men to die from heart disease as a result of strokes. A 50% increase in men's mortality was reported by the World Health Organization in 2009. The consequences of cardiovascular disease are severe. The causes of heart disease include diabetes, high blood pressure, high cholesterol, abnormal pulse rates, etc. Machine learning (ML) can be used to make predictions and decisions in the healthcare industry. Thus, scientists have turned to modern technologies like Machine Learning and Data Mining to predict diseases. The disease prediction is based on four algorithms. Compared to other boosts, the Ada boost is much more accurate.

Keywords: heart disease, cardiovascular disease, coronary artery disease, feature selection, random forest, AdaBoost, SVM, decision tree

Procedia PDF Downloads 155

Authors: Talar Agopian

Abstract:

Teaching English as a Second Language (ESL) is a common practice in many Lebanese schools. However, ESL teaching is done in traditional ways. Methods such as constructivism are seldom used, especially in villages. Here lies the significance of this research which joins constructivism and Piaget’s theory of cognitive development in ESL classes in Lebanese villages. The purpose of the present study is to explore the effects of applying constructivist student-centered strategies in teaching grammar, reading comprehension, and poetry on students in elementary ESL classes in two villages in Lebanon, Zefta in South Lebanon and Boqaata in Mount Lebanon. 20 English teachers participated in a training titled “Teaching English with Fun”, which focused on strategies that create a student-centered class where active learning takes place and there is increased learner engagement and autonomy. The training covered three main areas in teaching English: grammar, reading comprehension, and poetry. After participating in the training, the teachers applied the new strategies and methods in their ESL classes. The methodology comprised two phases: in phase one, practice-based research was conducted as the teachers attended the training and applied the constructivist strategies in their respective ESL classes. Phase two included the reflections of the teachers on the effects of the application of constructivist strategies. The results revealed the educational benefits of constructivist student-centered strategies; the students of teachers who applied these strategies showed improved engagement, positive attitudes towards poetry, increased motivation, and a better sense of autonomy. Future research is required in applying constructivist methods in the areas of writing, spelling, and vocabulary in ESL classrooms of Lebanese villages.

Keywords: active learning, constructivism, learner engagement, student-centered strategies

Procedia PDF Downloads 144

Authors: Thomas Q. Pan, Anika Basu, Chamith S. Rajapakse

Abstract:

Machine learning requires data that is categorized into features that models train on. This topic is important to the field of sports medicine due to the many tools it provides to physicians such as diagnosis support and risk assessment. Physician note that healthcare professionals take are usually unclean and not suitable for model training. The objective of this study was to develop and evaluate an advanced approach for extracting key features from sports medicine data without the need for extensive model training or data labeling. An LLM (Large Language Model) was given a narrative (Physician’s Notes) and prompted to extract four features (details about the patient). The narrative was found in a datasheet that contained six columns: Case Number, Validation Age, Validation Gender, Validation Diagnosis, Validation Body Part, and Narrative. The validation columns represent the accurate responses that the LLM attempts to output. With the given narrative, the LLM would output its response and extract the age, gender, diagnosis, and injured body part with each category taking up one line. The output would then be cleaned, matched, and added to new columns containing the extracted responses. Five ways of checking the accuracy were used: unclear count, substring comparison, LLM comparison, LLM re-check, and hand-evaluation. The unclear count essentially represented the extractions the LLM missed. This can be also understood as the recall score ([total - false negatives] over total). The rest of these correspond to the precision score ([total - false positives] over total). Substring comparison evaluated the validation (X) and extracted (Y) columns’ likeness by checking if X’s results were a substring of Y's findings and vice versa. LLM comparison directly asked an LLM if the X and Y’s results were similar. LLM Re-check prompted the LLM to see if the extracted results can be found in the narrative. Lastly, A selection of 1,000 random narratives was also selected and hand-evaluated to give an estimate of how well the LLM-based feature extraction model performed. With a selection of 10,000 narratives, the LLM-based approach had a recall score of roughly 98%. However, the precision scores of the substring comparison and LLM comparison models were around 72% and 76% respectively. The reason for these low figures is due to the minute differences between answers. For example, the ‘chest’ is a part of the ‘upper trunk’ however, these models cannot detect that. On the other hand, the LLM re-check and subset of hand-tested narratives showed a precision score of 96% and 95%. If this subset is used to extrapolate the possible outcome of the whole 10,000 narratives, the LLM-based approach would be strong in both precision and recall. These results indicated that an LLM-based feature extraction model could be a useful way for medical data in sports to be collected and analyzed by machine learning models. Wide use of this method could potentially increase the availability of data thus improving machine learning algorithms and supporting doctors with more enhanced tools.

Keywords: AI, LLM, ML, sports

Procedia PDF Downloads 12

Authors: Magaji Zainab Musa, Nordin M. A. Rahman, Julaily Aida Jusoh

Abstract:

This paper discusses the formal specification of web services applications for digital reference services (WSDRS). Digital reference service involves a user requesting for help from a reference librarian and a reference librarian responding to the request of a user all by electronic means. In most cases users do not get satisfied while using digital reference service due to delay of response of the librarians. Another may be due to no response or due to librarian giving an irrelevant solution to the problem submitted by the user. WDSRS is an informal model that claims to reduce the problems of digital reference services in libraries. It uses web services technology to provide efficient way of satisfying users’ need in the reference section of libraries. But informal model is in natural language which is inconsistent and ambiguous that may cause difficulties to the developers of the system. In order to solve this problem we decided to convert the informal specifications into formal specifications. This is supposed to reduce the overall development time and cost. Formal specification can be used to provide an unambiguous and precise supplement to natural language descriptions. It can be rigorously validated and verified leading to the early detection of specification errors. We use Z language to develop the formal model and verify it with Z/EVES theorem prover tool.

Keywords: formal, specifications, web services, digital reference services

Procedia PDF Downloads 377

Authors: Nishant Rodrigues, Nicole Spanedda, Chilukuri K. Mohan, Arindam Chakraborty

Abstract:

A crucial objective in quantum chemistry is the computation of the energy levels of chemical systems. This task requires electron-repulsion integrals as inputs, and the steep computational cost of evaluating these integrals poses a major numerical challenge in efficient implementation of quantum chemical software. This work presents a moment-based machine-learning approach for the efficient evaluation of electron-repulsion integrals. These integrals were approximated using linear combinations of a small number of moments. Machine learning algorithms were applied to estimate the coefficients in the linear combination. A random forest approach was used to identify promising features using a recursive feature elimination approach, which performed best for learning the sign of each coefficient but not the magnitude. A neural network with two hidden layers were then used to learn the coefficient magnitudes along with an iterative feature masking approach to perform input vector compression, identifying a small subset of orbitals whose coefficients are sufficient for the quantum state energy computation. Finally, a small ensemble of neural networks (with a median rule for decision fusion) was shown to improve results when compared to a single network.

Keywords: quantum energy calculations, atomic orbitals, electron-repulsion integrals, ensemble machine learning, random forests, neural networks, feature extraction

Procedia PDF Downloads 117

Authors: Fu-Chi Chuang, Yu-Liang, Chang, Wen-Der Wang

Abstract:

Our environment has been contaminated by many artificial chemicals, such as plastics, pesticides. Many of them have hormone-like activity and are classified as 'environmental hormone (also named endocrine disruptors)'. These chemicals interfere with or mimic hormones have adverse effects that persist into adulthood. Environmental education is an important way to teach students to become engaged in real-world issues that transcend classroom walls. Elementary education is the first stage to perform environmental education and it is an important component to help students develop adequate environmental knowledge, attitudes, and behavior. However, elementary teachers' knowledge plays a critical role in this mission. Therefore, we use a questionnaire to survey the knowledge of environmental hormone of elementary school teachers and their learning motivation of the environmental hormone-regarding knowledge. We collected 218 questionnaires from Taiwanese elementary teachers and the results indicate around 73% of elementary teachers do not have enough knowledge about environmental hormones. Our results also reveal the in-service elementary teachers’ learning motivation of environmental hormones knowledge is positively enhanced once they realized their insufficient cognitive ability of environmental hormones. We believe our study will provide the powerful reference for Ministry of Education to set up the policy of environmental education to enrich all citizens sufficient knowledge of the effects of the environmental hormone on organisms, and further to enhance our correct environmental behaviors.

Keywords: elementary teacher, environmental hormones, learning motivation, questionnaire

Procedia PDF Downloads 314

Authors: Abdelaadim Bidaoui

Abstract:

As one of the most important linguistic inquiries, the topic of the intertwined relationship between language, the mind, and the world has attracted many scholars. In the fifties, Sapir and Whorf advocated the hypothesis that language shapes our cultural realities as an early attempt to provide answers to this linguistic inquiry. Later, discursive psychology views the linguistic behavior as “a dynamic form of social practice which constructs the social world, individual selves and identity.” (Jorgensen & Phillips 2002, 118). Discursive psychology also considers discourse as a trigger of social action and change. Building on discursive psychology and motivation theory, this paper examines the impact of linguistic behavior of the Moroccan coach Walid Reggragui on the Moroccan team’s exceptional performance in Qatar 2022 Soccer World Cup. The data used in the research is based on interviews conducted by the Moroccan coach prior and during the World Cup. Using a discourse analysis of the linguistic behavior of Reggragui, this paper shows how the linguistic behavior of Reggragui provided support for the three psychological needs: sense of belonging, competence, and autonomy. As any CDA research, this paper uses a triangulated theoretical framework that includes language, cognition and society.

Keywords: critical discourse analysis, motivation theory, discursive psychology, linguistic behavior

Procedia PDF Downloads 91

Authors: Ilona Buchem, Carolin Gellner

Abstract:

Over the last few years, the call for the support of senior learners in the development of their digital literacy has become prevalent, mainly due to the progression towards ageing societies paired with advances in digitalisation in all spheres of life, including e-health and electronic patient record (EPA). While major research efforts in supporting senior learners in developing digital literacy have been invested so far in e-learning focusing on knowledge acquisition and cognitive tasks, little research exists in learning models which target virtual mentoring and coaching with the help of pedagogical agents and address the social dimensions of learning. Research from studies with students in the context of formal education has already provided methods for designing intelligent virtual agents in support of personalised learning. However, this research has mostly focused on cognitive skills and has not yet been applied to the context of mentoring/coaching of senior learners, who have different characteristics and learn in different contexts. In this paper, we describe how insights from previous research can be used to develop an intelligent virtual learning coach (agent) for senior learners with a focus on building the social relationship between the agent and the learner and the key task of the agent to socialize learners to the larger context of digital literacy with a focus on electronic health records. Following current approaches to mentoring and coaching, the agent is designed not to enhance and monitor the cognitive performance of the learner but to serve as a trusted friend and advisor, whose role is to provide one-to-one guidance and support sharing of experiences among learners (peers). Based on literature review and synopsis of research on virtual agents and current coaching/mentoring models under consideration of the specific characteristics and requirements of senior learners, we describe the design framework which was applied to design an intelligent virtual learning coach as part of the e-learning system for digital literacy of senior learners in the ePA-Coach project founded by the German Ministry of Education and Research. This paper also presents the results from the evaluation study, which compared the use of the first prototype of the virtual learning coach designed according to the design framework with a voice narration in a multimedia learning environment with senior learners. The focus of the study was to validate the agent design in the context of the persona effect (Lester et al., 1997). Since the persona effect is related to the hypothesis that animated agents are perceived as more socially engaging, the study evaluated possible impacts of agent coaching in comparison with voice coaching on motivation, engagement, experience, and digital literacy.

Keywords: virtual learning coach, virtual mentor, pedagogical agent, senior learners, digital literacy, electronic health records

Procedia PDF Downloads 118

Authors: Sachin Nagargoje

Abstract:

Complete labeled data is often difficult to obtain in a practical scenario. Even if one manages to obtain the data, the quality of the data is always in question. In shopping vertical, offers are the input data, which is given by advertiser with or without a good quality of information. In this paper, an author investigated the possibility of using a very simple Semi-supervised learning approach to increase the recall of unhealthy offers (has badly written Offer Title or partial product details) in shopping vertical domain. The author found that the semisupervised learning method had improved the recall in the Smart Phone category by 30% on A=B testing on 10% traffic and increased the YoY (Year over Year) number of impressions per month by 33% at production. This also made a significant increase in Revenue, but that cannot be publicly disclosed.

Keywords: semi-supervised learning, clustering, recall, coverage

Procedia PDF Downloads 123

Authors: Eunice H. Li

Abstract:

This e-poster explores some of the issues that are linked to Open Educational Resources (OER). It describes how OER is explained by experts in the field and relates its value in attaining and using knowledge. ‘Open', 'open pedagogy', self-direction, freedom, and autonomy are the main issues identified for the discussion. All of these issues make essential contributions to OER in one way or another. Nevertheless, there are seemingly areas of contentions with regard to applying these concepts in teaching and learning practices. For this e-Poster, it is the teaching-learning aspects of OER that it is primarily concerned with. The basis for the discussion comes from a 2013 critique of OER presented by Jeremy Knox of the University of Edinburgh, tutor of the MSc in Digital Education Programme. This discussion is also supported by the analysis of other research work and papers in this area. The general view on OER is that it is a useful tool for the advancement of learner-centred models of education, but in whatever context, pedagogy cannot be diminished and overlooked. It should take into consideration how to deal with the issues identified above in order to allow learners to gain full benefit from OER.

Keywords: open, pedagogy, e-learning technologies, autonomy, knowledge

Procedia PDF Downloads 401

Authors: Daniel Asiamah Ameyaw

Abstract:

This research attempts to investigate how lecturer evaluation makes a difference in enhancing higher education teaching and learning processes. The research questions to guide this research work states first as, “What are the perspectives on the difference made by evaluating academic teachers in order to enhance higher education teaching and learning processes?” and second, “What are the implications of the findings for Policy and Practice?” Data for this research was collected mainly through interviewing and partly documents review. Data analysis was conducted under the framework of grounded theory. The findings showed that for individual lecturer level, lecturer evaluation provides a continuous improvement of teaching strategies, and serves as source of data for research on teaching. At the individual student level, it enhances students learning process; serving as source of information for course selection by students; and by making students feel recognised in the educational process. At the institutional level, it noted that lecturer evaluation is useful in personnel and management decision making; it assures stakeholders of quality teaching and learning by setting up standards for lecturers; and it enables institutions to identify skill requirement and needs as a basis for organising workshops. Lecturer evaluation is useful at national level in terms of guaranteeing the competencies of graduates who then provide the needed manpower requirement of the nation. Besides, it mentioned that resource allocation to higher educational institution is based largely on quality of the programmes being run by the institution. The researcher concluded, that the findings have implications for policy and practice, therefore, higher education managers are expected to ensure that policy is implemented as planned by policy-makers so that the objectives can successfully be achieved.

Keywords: academic quality, higher education, lecturer evaluation, teaching and learning processes

Procedia PDF Downloads 146

Authors: Azusa Katsumata

Abstract:

Where there is collaboration between universities and companies in the educational context, companies seek ‘knowledge’ from universities and provide a ‘place of practice’ to them. Several universities have introduced activities aimed at the mutual enlightenment of a diversity of people in career education. However, several programs emphasize on delivering results, and on practicing the prepared materials as planned. Few programs focus on unexpected failures and setbacks. This way of learning is important in career education so that classmates can help each other, overcome difficulties, draw out each other’s strengths, and learn from them. Seijo University in Tokyo offered Tokyo Tourism, a Project-Based Learning course, as a first-year career education course until 2016. In cooperation with a travel agency, students participate in planning actual tourism products for foreigners visiting Japan, undertake tours serving as guides. This paper aims to study the 'learning platform' created by a series of processes such as the fieldwork, planning tours, the presentation, selling the tourism products, and guiding the tourists. We conducted a questionnaire to measure the development of work-related skills in class. From the results of the questionnaire, we can see, in the example of this class, that students demonstrated an increased desire to be pro-active and an improved motivation to learn. Students have not, however, acquired policy or business skills. This is appropriate for first-year careers education, but we need to consider how this can be incorporated into future courses. In the questionnaire filled out by the students after the class, the following results were found. Planning and implementing travel products while learning from each other, and helping the teams has led to improvements in the student workforce. This course is a collaborative project between Japanese universities and the 2020 Tokyo Olympics and Paralympic Games committee.

Keywords: university career education, platform of learning, project-based learning, collaboration between university and company

Procedia PDF Downloads 163

Authors: Walid Cherif

Abstract:

Data mining has, over recent years, seen big advances because of the spread of internet, which generates everyday a tremendous volume of data, and also the immense advances in technologies which facilitate the analysis of these data. In particular, classification techniques are a subdomain of Data Mining which determines in which group each data instance is related within a given dataset. It is used to classify data into different classes according to desired criteria. Generally, a classification technique is either statistical or machine learning. Each type of these techniques has its own limits. Nowadays, current data are becoming increasingly heterogeneous; consequently, current classification techniques are encountering many difficulties. This paper defines new measure functions to quantify the resemblance between instances and then combines them in a new approach which is different from actual algorithms by its reliability computations. Results of the proposed approach exceeded most common classification techniques with an f-measure exceeding 97% on the IRIS Dataset.

Keywords: data mining, knowledge discovery, machine learning, similarity measurement, supervised classification

Procedia PDF Downloads 465