Search results for: teaching learning model

Authors: Maha Wiss, Wael Khreich

Abstract:

Animals, objects, foods, plants, and other non-human terms are commonly used as a source of metaphors to describe females in formal and slang language. Comparing women to non-human items not only reflects cultural views that might conceptualize women as subordinates or in a lower position than humans, yet it conveys this degradation to the listeners. Moreover, the dehumanizing representation of females in the language normalizes the derogation and even encourages sexism and aggressiveness against women. Although dehumanization has been a popular research topic for decades, according to our knowledge, no studies have linked women's dehumanizing language to the machine learning field. Therefore, we introduce our research work as one of the first attempts to create a tool for the automated detection of the dehumanizing depiction of females in English texts. We also present the first labeled dataset on the charted topic, which is used for training supervised machine learning algorithms to build an accurate classification model. The importance of this work is that it accomplishes the first step toward mitigating dehumanizing language against females.

Keywords: gender bias, machine learning, NLP, women dehumanization

Procedia PDF Downloads 80

Authors: R. O. Osaseri, A. R. Usiobaifo

Abstract:

The power transformer which is responsible for the voltage transformation is of great relevance in the power system and oil-immerse transformer is widely used all over the world. A prompt and proper maintenance of the transformer is of utmost importance. The dissolved gasses content in power transformer, oil is of enormous importance in detecting incipient fault of the transformer. There is a need for accurate prediction of the incipient fault in transformer oil in order to facilitate the prompt maintenance and reducing the cost and error minimization. Study on fault prediction and diagnostic has been the center of many researchers and many previous works have been reported on the use of artificial intelligence to predict incipient failure of transformer faults. In this study machine learning technique was employed by using gradient decent algorithms and Support Vector Machine (SVM) in predicting incipient fault diagnosis of transformer. The method focuses on creating a system that improves its performance on previous result and historical data. The system design approach is basically in two phases; training and testing phase. The gradient decent algorithm is trained with a training dataset while the learned algorithm is applied to a set of new data. This two dataset is used to prove the accuracy of the proposed model. In this study a transformer fault diagnostic model based on Support Vector Machine (SVM) and gradient decent algorithms has been presented with a satisfactory diagnostic capability with high percentage in predicting incipient failure of transformer faults than existing diagnostic methods.

Keywords: diagnostic model, gradient decent, machine learning, support vector machine (SVM), transformer fault

Procedia PDF Downloads 324

Authors: Abdullah Abu-Tineh, Carol Murphy, Nigel Calder, Nasser Mansour

Abstract:

This paper reports on an aspect of e-learning in developing inquiry-based learning (IBL). We present data on the views of teachers and students in Qatar following a professional development programme intended to help teachers implement IBL in their science and mathematics classrooms. Key to this programme was the use of WebQuests. Views of the teachers and students suggested that WebQuests helped students to develop technical skills, work collaboratively and become independent in their learning. The use of WebQuests also enabled a combination of digital and non-digital tools that helped students connect ideas and enhance their understanding of topics.

Keywords: digital technology, inquiry-based learning, mathematics and science education, professional development

Procedia PDF Downloads 143

Authors: Raluca Ionela Maxim

Abstract:

Designing a gender responsive Science, Technology, Engineering and Mathematics (STEM) mobile game based learning solution (mGBL) is a challenge in terms of content, gamification level and equal engagement of girls and boys. The goal of this case study was to research and create a high-fidelity prototype design of a mobile game that contains role-models as avatars that guide and expose girls and boys to STEM learning content. For this research purpose it was applied the methodology of design sprint with five-phase process that combines design thinking principles. The technique of this methodology comprises smart interviews with STEM experts, mind-map creation, sketching, prototyping and usability testing of the interactive prototype of the gender responsive STEM mGBL. The results have shown that the effect of the avatar/role model had a positive impact. Therefore, by exposing students (boys and girls) to STEM role models in an mGBL tool is helpful for the decreasing of the gender inequalities in STEM fields.

Keywords: design thinking, design sprint, gender-responsive STEM education, mobile game based learning, role-models

Procedia PDF Downloads 135

Authors: Oumaima Khlifati, Khadija Baba

Abstract:

Pavement distress is the main factor responsible for the deterioration of road structure durability, damage vehicles, and driver comfort. Transportation agencies spend a high proportion of their funds on pavement monitoring and maintenance. The auscultation of pavement distress was based on the manual survey, which was extremely time consuming, labor intensive, and required domain expertise. Therefore, the automatic distress detection is needed to reduce the cost of manual inspection and avoid more serious damage by implementing the appropriate remediation actions at the right time. Inspired by recent deep learning applications, this paper proposes an algorithm for automatic road distress detection and classification using on the Deep Convolutional Neural Network (DCNN). In this study, the types of pavement distress are classified as transverse or longitudinal cracking, alligator, pothole, and intact pavement. The dataset used in this work is composed of public asphalt pavement images. In order to learn the structure of the different type of distress, the DCNN models are trained and tested as a multi-label classification task. In addition, to get the highest accuracy for our model, we adjust the structural optimization hyper parameters such as the number of convolutions and max pooling, filers, size of filters, loss functions, activation functions, and optimizer and fine-tuning hyper parameters that conclude batch size and learning rate. The optimization of the model is executed by checking all feasible combinations and selecting the best performing one. The model, after being optimized, performance metrics is calculated, which describe the training and validation accuracies, precision, recall, and F1 score.

Keywords: distress pavement, hyperparameters, automatic classification, deep learning

Procedia PDF Downloads 93

Authors: B. R. Campomanes-Alvarez, P. Quiros, B. Fernandez

Abstract:

Automatic Speech Recognition (ASR) is a machine-based process of decoding and transcribing oral speech. A typical ASR system receives acoustic input from a speaker or an audio file, analyzes it using algorithms, and produces an output in the form of a text. Some speech recognition systems use Hidden Markov Models (HMMs) to deal with the temporal variability of speech and Gaussian Mixture Models (GMMs) to determine how well each state of each HMM fits a short window of frames of coefficients that represents the acoustic input. Another way to evaluate the fit is to use a feed-forward neural network that takes several frames of coefficients as input and produces posterior probabilities over HMM states as output. Deep neural networks (DNNs) that have many hidden layers and are trained using new methods have been shown to outperform GMMs on a variety of speech recognition systems. Acoustic models for state-of-the-art ASR systems are usually training on massive amounts of data. However, audio files with their corresponding transcriptions can be difficult to obtain, especially in the Spanish language. Hence, in the case of these low-resource scenarios, building an ASR model is considered as a complex task due to the lack of labeled data, resulting in an under-trained system. Semi-supervised learning approaches arise as necessary tasks given the high cost of transcribing audio data. The main goal of this proposal is to develop a procedure based on acoustic semi-supervised learning for Spanish ASR systems by using DNNs. This semi-supervised learning approach consists of: (a) Training a seed ASR model with a DNN using a set of audios and their respective transcriptions. A DNN with a one-hidden-layer network was initialized; increasing the number of hidden layers in training, to a five. A refinement, which consisted of the weight matrix plus bias term and a Stochastic Gradient Descent (SGD) training were also performed. The objective function was the cross-entropy criterion. (b) Decoding/testing a set of unlabeled data with the obtained seed model. (c) Selecting a suitable subset of the validated data to retrain the seed model, thereby improving its performance on the target test set. To choose the most precise transcriptions, three confidence scores or metrics, regarding the lattice concept (based on the graph cost, the acoustic cost and a combination of both), was performed as selection technique. The performance of the ASR system will be calculated by means of the Word Error Rate (WER). The test dataset was renewed in order to extract the new transcriptions added to the training dataset. Some experiments were carried out in order to select the best ASR results. A comparison between a GMM-based model without retraining and the DNN proposed system was also made under the same conditions. Results showed that the semi-supervised ASR-model based on DNNs outperformed the GMM-model, in terms of WER, in all tested cases. The best result obtained an improvement of 6% relative WER. Hence, these promising results suggest that the proposed technique could be suitable for building ASR models in low-resource environments.

Keywords: automatic speech recognition, deep neural networks, machine learning, semi-supervised learning

Procedia PDF Downloads 340

Authors: Sadia Nasrin Tisha, Mushfika Sharmin Rahman, Javier Orduz

Abstract:

Machine learning is applied in a variety of fields throughout the world. The healthcare sector has benefited enormously from it. One of the most effective approaches for predicting human heart diseases is to use machine learning applications to classify data and predict the outcome as a classification. However, with the rapid advancement of quantum technology, quantum computing has emerged as a potential game-changer for many applications. Quantum algorithms have the potential to execute substantially faster than their classical equivalents, which can lead to significant improvements in computational performance and efficiency. In this study, we applied quantum machine learning concepts to predict coronary heart diseases from text data. We experimented thrice with three different features; and three feature sets. The data set consisted of 100 data points. We pursue to do a comparative analysis of the two approaches, highlighting the potential benefits of quantum machine learning for predicting heart diseases.

Keywords: quantum machine learning, SVM, QSVM, matrix product state

Procedia PDF Downloads 94

Authors: Wilma Coetzee

Abstract:

Statistics lecturers experience that many students who are taking a service course in statistics do not like statistics. Students in these courses tend to struggle and do not perform well. This research takes a look at the student’s perspective, with the aim to determine how to change the teaching of statistics so that students will enjoy it more and perform better. Questionnaires were used to determine the perspectives of first year service statistics students at a South African university. Factors addressed included motivation to study, attitude toward statistics, statistical anxiety, mathematical abilities and tendency to procrastinate. Logistic regression was used to determine what contributes to students performing badly in statistics. The results show that the factors that contribute the most to students performing badly are: statistical anxiety, not being motivated and having had mathematical literacy instead of mathematics in secondary school. Two open ended questions were included in the questionnaire: 'I will enjoy statistics more if…' and 'I will perform better in statistics if…'. The answers to these questions were analyzed using qualitative methods. Frequent themes were identified for each of the questions. A simulation study incorporating bootstrapping was done to determine the saturation of the themes. The majority of the students indicated that they would perform better in statistics if they studied more, managed their time better, had a flare for mathematics and if the lecturer was able to explain difficult concepts better. They also want more active learning. To ensure that students enjoy statistics more, they want an active learning experience. They want fun activities, more interaction with the lecturer and with one another, more computer based problems, and more challenges. They want a better understanding of the subject, want to understand the relevance of statistics to their future career and want excellent lecturers. These findings can be used to direct the improvement of the tuition of statistics.

Keywords: active learning, performance in statistics, statistical anxiety, statistics education

Procedia PDF Downloads 147

Authors: Souhir Ben Amor, Heni Boubaker, Lotfi Belkacem

Abstract:

This aims of this paper is to forecast the electricity spot prices. First, we focus on modeling the conditional mean of the series so we adopt a generalized fractional -factor Gegenbauer process (k-factor GARMA). Secondly, the residual from the -factor GARMA model has used as a proxy for the conditional variance; these residuals were predicted using two different approaches. In the first approach, a local linear wavelet neural network model (LLWNN) has developed to predict the conditional variance using the Back Propagation learning algorithms. In the second approach, the Gegenbauer generalized autoregressive conditional heteroscedasticity process (G-GARCH) has adopted, and the parameters of the k-factor GARMA-G-GARCH model has estimated using the wavelet methodology based on the discrete wavelet packet transform (DWPT) approach. The empirical results have shown that the k-factor GARMA-G-GARCH model outperform the hybrid k-factor GARMA-LLWNN model, and find it is more appropriate for forecasts.

Keywords: electricity price, k-factor GARMA, LLWNN, G-GARCH, forecasting

Procedia PDF Downloads 231

Authors: Abigaile Rose Mary R. Capay, Janne Ly Castillon-Gilpo, Sheila A. Javier

Abstract:

The transition to online learning has been a challenging feat on the mental health of tertiary students. This study investigated whether learning mindfulness strategies online would help in improving students’ imagination, conscientiousness, extraversion, agreeableness and emotional stability, as measured by the International Personality Item Pool (IPIP) Big Five Factor Markers, as well as their dispositional mindfulness as measured by the Mindfulness Attention Awareness Scale (MAAS). Fifty-two college students participated in the experiment. The 23 participants assigned to the treatment condition received 6-weekly experiential sessions of online mindfulness training and were advised to follow a daily mindfulness practice, while the 29 participants from the control group only received a 1-hour lecture. Scores were collected at pretest and posttest. Findings show that there was a significant difference in the pretest and posttest scores of students assigned in the treatment group, likewise medium effect sizes in the variables: dispositional mindfulness (t (22) = 2.64, p = 0.015, d = .550), extraversion (t (22) = 2.76, p = 0.011, d = 0.575), emotional stability (t (22) = 2.99, p = 0.007, d = .624), conscientiousness (t (22) = 2.74, p = 0.012, d = .572) and imagination (t (22) = 4.08, p < .001), but not for agreeableness (t (22) = 2.01, p = 0.057, d = .419). No significant differences were observed on the scores of the control group. Educational institutions are recommended to consider teaching basic mindfulness strategies to tertiary students, as a valuable resource in improving their mental health as they navigate through adjustments in online learning.

Keywords: mindfulness, school-based interventions, MAAS, IPIP Big Five Markers, experiment

Procedia PDF Downloads 58

Authors: Charlotte Brenner

Abstract:

Teachers’ beleifs about inclusion are foundational to their implementation of inclusive teaching practices. Utilizing a longitudinal design and multiple case study methodology, this study investigates how teacher candidates’ instructional and practicum experiences shape their beliefs about inclusion in one teacher education program located in western Canada (N=20). Interview questions were developed through the lens of self-determinaiton theory and theory about teachers’ beleifs and inclusion. Preliminary thematic ananysis indicates that a 36-hour course focused on diversity and inclusion supports teacher candiates to deepen their understandings of: the need for inclusion in classrooms and strategies to promote inclusion. Furthermore, teacher candiates identified course components that fostered their developing understandings of inclusion. Future data will examine the stability of teacher candidates’ beliefs about inclusion and their implementation of inclusive teaching strategies throughout their practicum experiences.

Keywords: teacher candidates, inclusion, teacher education programs, beliefs

Procedia PDF Downloads 88

Authors: Christian Neunzig, Simon Fahle, Jürgen Schulz, Matthias Möller, Bernd Kuhlenkötter

Abstract:

Due to increasing cost pressure in global markets, artificial intelligence is becoming a technology that is decisive for competition. Predictive quality enables machinery and plant manufacturers to ensure product quality by using data-driven forecasts via machine learning models as a decision-making basis for test results. The use of cross-process Bosch production data along the value chain of hydraulic valves is a promising approach to classifying the quality characteristics of workpieces.

Keywords: predictive quality, hydraulics, machine learning, classification, supervised learning

Procedia PDF Downloads 231

Authors: C. E. Laburu, M. A. Barros, A. F. Zompero, O. H. M. Silva

Abstract:

The work presents a teaching strategy that employs syntagmatic and paradigmatic linguistic relations in order to monitor the understanding of physics students’ concepts. Syntagmatic and paradigmatic relations are theoretical elements of semiotics studies and our research circumstances and justified them within the research program of multi-modal representations. Among the multi-modal representations to learning scientific knowledge, the scope of action of syntagmatic and paradigmatic relations belongs to the discursive writing form. The use of such relations has the purpose to seek innovate didactic work with discourse representation in the write form before translate to another different representational form. The research was conducted with a sample of first year high school students. The students were asked to produce syntagmatic and paradigmatic of Newton’ first law statement. This statement was delivered in paper for each student that should individually write the relations. The student’s records were collected for analysis. It was possible observed in one student used here as example that their monemes replaced and rearrangements produced by, respectively, syntagmatic and paradigmatic relations, kept the original meaning of the law. In paradigmatic production he specified relevant significant units of the linguistic signs, the monemas, which constitute the first articulation and each word substituted kept equivalence to the original meaning of original monema. Also, it was noted a number of diverse and many monemas were chosen, with balanced combination of grammatical (grammatical monema is what changes the meaning of a word, in certain positions of the syntagma, along with a relatively small number of other monemes. It is the smallest linguistic unit that has grammatical meaning) and lexical (lexical monema is what belongs to unlimited inventories; is the monema endowed with lexical meaning) monemas. In syntagmatic production, monemas ordinations were syntactically coherent, being linked with semantic conservation and preserved number. In general, the results showed that the written representation mode based on linguistic relations paradigmatic and syntagmatic qualifies itself to be used in the classroom as a potential identifier and accompanist of meanings acquired from students in the process of scientific inquiry.

Keywords: semiotics, language, high school, physics teaching

Procedia PDF Downloads 131

Authors: Edgardo V. Gubatanga Jr., Mark Joshua Salvacion

Abstract:

Most aerial LiDAR systems have accompanying aerial cameras in order to capture not only the terrain of the surveyed area but also its true-color appearance. However, the presence of atmospheric clouds, poor lighting conditions, and aerial camera problems during an aerial survey may cause absence of aerial photographs. These leave areas having terrain information but lacking aerial photographs. Intensity images can be derived from LiDAR data but they are only grayscale images. A deep learning model is developed to create a complex function in a form of a deep neural network relating the pixel values of LiDAR-derived intensity images and true-color images. This complex function can then be used to predict the true-color images of a certain area using intensity images from LiDAR data. The predicted true-color images do not necessarily need to be accurate compared to the real world. They are only intended to look realistic so that they can be used as base maps.

Keywords: aerial LiDAR, colorization, deep learning, intensity images

Procedia PDF Downloads 166

Authors: Mohamad Abdullah

Abstract:

Recent Advancements in technology have stimulated deep change in the language learning classroom. Flipped classroom as a new pedagogical method is at the center of this change. It turns the classroom into a student-centered environment and promotes interactive and autonomous learning. The present study is an attempt to examine the effectiveness of the Flipped Classroom Model (FCM) on students’ motivation level in English speaking performance. This study was carried out with 27 undergraduate female English majors who enrolled in the course of Advanced Communication Skills (ENGL 154) at Buraimi University College (BUC). Data was collected through Motivation in English Speaking Performance Questionnaire (MESPQ) which has been distributed among the participants of this study pre and post the implementation of FCM. SPSS was used for analyzing data. The Paired T-Test which was carried out on the pre-post of (MESPQ) showed a significant difference between them (p < .009) that revealed participants’ tendency to increase their motivation level in English speaking performance after the application of FCM. In addition, respondents of the current study reported positive views about the implementation of FCM.

Keywords: english speaking performance, motivation, flipped classroom model, learner-contentedness

Procedia PDF Downloads 132

Authors: Nezha Badi

Abstract:

It has been shown in the literature that Algerian university students suffer from low levels of reading comprehension proficiency, which hinder their overall proficiency in English. This low level is mainly related to the methodology of teaching reading which is employed by the teacher in the classroom (a teacher-centered environment), as well as students’ poor sense of self-efficacy to undertake reading comprehension activities. Arguably, what is needed is an approach necessary for enhancing students’ self-beliefs about their abilities to deal with different reading comprehension activities. This can be done by providing them with opportunities to take responsibility for their own learning (learners’ autonomy). As a result of learning autonomy, learners’ beliefs about their abilities to deal with certain language tasks may increase, and hence, their language learning ability. Therefore, this experimental research study attempts to assess the extent to which an integrated approach combining one particular reading approach known as ‘collaborative strategic reading’ (CSR), and teacher’s attributional feedback (on students’ reading performance and strategy use) can improve the reading comprehension skill and the sense of self-efficacy of EFL Algerian university students. It also seeks to examine students’ main reasons for their successful or unsuccessful achievements in reading comprehension activities, and whether students’ attributions for their reading comprehension outcomes can be modified after exposure to the instruction. To obtain the data, different tools including a reading comprehension test, questionnaires, an observation, an interview, and learning logs were used with 105 second year Algerian EFL university students. The sample of the study was divided into three groups; one control group (with no treatment), one experimental group (CSR group) who received a CSR instruction, and a second intervention group (CSR Plus group) who received teacher’s attribution feedback in addition to the CSR intervention. Students in the CSR Plus group received the same experiment as the CSR group using the same tools, except that they were asked to keep learning logs, for which teacher’s feedback on reading performance and strategy use was provided. The results of this study indicate that the CSR and the attributional feedback intervention was effective in improving students’ reading comprehension proficiency and sense of self-efficacy. However, there was not a significant change in students’ adaptive and maladaptive attributions for their success and failure d from the pre-test to the post-test phase. Analysis of the perception questionnaire, the interview, and the learning logs shows that students have positive perceptions about the CSR and the attributional feedback instruction. Based on the findings, this study, therefore, seeks to provide EFL teachers in general and Algerian EFL university teachers in particular with pedagogical implications on how to teach reading comprehension to their students to help them achieve well and feel more self-efficacious in reading comprehension activities, and in English language learning more generally.

Keywords: attributions, attributional feedback, collaborative strategic reading, self-efficacy

Procedia PDF Downloads 119

Authors: M. M. Essa, S. Subash, M. Akbar, S. Al-Adawi, A. Al-Asmi, G. J. Guillemein

Abstract:

Objective: Transgenic (tg) mice which contain an amyloid precursor protein (APP) gene mutation, develop extracellular amyloid beta (Aβ) deposition in the brain, and severe memory and behavioural deficits with age. These mice serve as an important animal model for testing the efficacy of novel drug candidates for the treatment and management of symptoms of Alzheimer's disease (AD). Several reports have suggested that oxidative stress is the underlying cause of Aβ neurotoxicity in AD. Pomegranates contain very high levels of antioxidants and several medicinal properties that may be useful for improving the quality of life in AD patients. In this study, we investigated the effect of dietary supplementation of Omani pomegranate extract on the memory, anxiety and learning skills along with inflammation in an AD mouse model containing the double Swedish APP mutation (APPsw/Tg2576). Methods: The experimental groups of APP-transgenic mice from the age of 4 months were fed custom-mix diets (pellets) containing 4% pomegranate. We assessed spatial memory and learning ability, psychomotor coordination, and anxiety-related behavior in Tg and wild-type mice at the age of 4-5 months and 18-19 months using the Morris water maze test, rota rod test, elevated plus maze test, and open field test. Further, inflammatory parameters also analysed. Results: APPsw/Tg2576 mice that were fed a standard chow diet without pomegranates showed significant memory deficits, increased anxiety-related behavior, and severe impairment in spatial learning ability, position discrimination learning ability and motor coordination along with increased inflammation compared to the wild type mice on the same diet, at the age of 18-19 months In contrast, APPsw/Tg2576 mice that were fed a diet containing 4% pomegranates showed a significant improvements in memory, learning, locomotor function, and anxiety with reduced inflammatory markers compared to APPsw/Tg2576 mice fed the standard chow diet. Conclusion: Our results suggest that dietary supplementation with pomegranates may slow the progression of cognitive and behavioural impairments in AD. The exact mechanism is still unclear and further extensive research needed.

Keywords: Alzheimer's disease, pomegranates, oman, cognitive decline, memory loss, anxiety, inflammation

Procedia PDF Downloads 529

Authors: Rabi Abdullahi Danjuma, Fatima Binta Attahir

Abstract:

Classroom discourse is important, and it is worth examining what the phenomena is and how it helps both the teacher and students in a classroom situation. This paper looks at the classroom as a traditional social setting which has its own norms and values. The paper also explains what discourse is, as extended communication in speech or writing often interactively dealing with some particular topics. It also discusses classroom discourse as the language which teachers and students use to communicate with each other in a classroom situation. The paper also looks at some strategies for effective classroom discourse. Finally, implications and recommendations were drawn.

Keywords: classroom, discourse, learning, student, strategies, communication

Procedia PDF Downloads 607

Authors: Yash Jain

Abstract:

The US Center for Disease Control has recommended wearing masks to slow the spread of the virus. The research uses a video feed from a camera to conduct real-time classifications of whether or not a human is correctly wearing a mask, incorrectly wearing a mask, or not wearing a mask at all. Utilizing two distinct datasets from the open-source website Kaggle, a mask detection network had been trained. The first dataset that was used to train the model was titled 'Face Mask Detection' on Kaggle, where the dataset was retrieved from and the second dataset was titled 'Face Mask Dataset, which provided the data in a (YOLO Format)' so that the TinyYoloV3 model could be trained. Based on the data from Kaggle, two machine learning models were implemented and trained: a Tiny YoloV3 Real-time model and a two-stage neural network classifier. The two-stage neural network classifier had a first step of identifying distinct faces within the image, and the second step was a classifier to detect the state of the mask on the face and whether it was worn correctly, incorrectly, or no mask at all. The TinyYoloV3 was used for the live feed as well as for a comparison standpoint against the previous two-stage classifier and was trained using the darknet neural network framework. The two-stage classifier attained a mean average precision (MAP) of 80%, while the model trained using TinyYoloV3 real-time detection had a mean average precision (MAP) of 59%. Overall, both models were able to correctly classify stages/scenarios of no mask, mask, and incorrectly worn masks.

Keywords: datasets, classifier, mask-detection, real-time, TinyYoloV3, two-stage neural network classifier

Procedia PDF Downloads 163

Authors: Pantaleon Lutta, Mohamed Sedky, Mohamed Hassan

Abstract:

The inability of organizations to put in place management control measures for Internet of Things (IoT) complexities persists to be a risk concern. Policy makers have been left to scamper in finding measures to combat these security and privacy concerns. IoT forensics is a cumbersome process as there is no standardization of the IoT products, no or limited historical data are stored on the devices. This paper highlights why IoT forensics is a unique adventure and brought out the legal challenges encountered in the investigation process. A quadrant model is presented to study the conflicting aspects in IoT forensics. The model analyses the effectiveness of forensic investigation process versus the admissibility of the evidence integrity; taking into account the user privacy and the providers’ compliance with the laws and regulations. Our analysis concludes that a semi-automated forensic process using machine learning, could eliminate the human factor from the profiling and surveillance processes, and hence resolves the issues of data protection (privacy and confidentiality).

Keywords: cloud forensics, data protection Laws, GDPR, IoT forensics, machine Learning

Procedia PDF Downloads 150

Authors: Shrabani Ghosh

Abstract:

Automatic offensive language detection in social media has become a stirring task in today's NLP. Manual Offensive language detection is tedious and laborious work where automatic methods based on machine learning are only alternatives. Previous works have done sentiment analysis over social media in different ways such as supervised, semi-supervised, and unsupervised manner. Domain adaptation in a semi-supervised way has also been explored in NLP, where the source domain and the target domain are different. In domain adaptation, the source domain usually has a large amount of labeled data, while only a limited amount of labeled data is available in the target domain. Pretrained transformers like BERT, RoBERTa models are fine-tuned to perform text classification in an unsupervised manner to perform further pre-train masked language modeling (MLM) tasks. In previous work, hate speech detection has been explored in Gab.ai, which is a free speech platform described as a platform of extremist in varying degrees in online social media. In domain adaptation process, Twitter data is used as the source domain, and Gab data is used as the target domain. The performance of domain adaptation also depends on the cross-domain similarity. Different distance measure methods such as L2 distance, cosine distance, Maximum Mean Discrepancy (MMD), Fisher Linear Discriminant (FLD), and CORAL have been used to estimate domain similarity. Certainly, in-domain distances are small, and between-domain distances are expected to be large. The previous work finding shows that pretrain masked language model (MLM) fine-tuned with a mixture of posts of source and target domain gives higher accuracy. However, in-domain performance of the hate classifier on Twitter data accuracy is 71.78%, and out-of-domain performance of the hate classifier on Gab data goes down to 56.53%. Recently self-supervised learning got a lot of attention as it is more applicable when labeled data are scarce. Few works have already been explored to apply self-supervised learning on NLP tasks such as sentiment classification. Self-supervised language representation model ALBERTA focuses on modeling inter-sentence coherence and helps downstream tasks with multi-sentence inputs. Self-supervised attention learning approach shows better performance as it exploits extracted context word in the training process. In this work, a self-supervised attention mechanism has been proposed to detect hate speech on Gab.ai. This framework initially classifies the Gab dataset in an attention-based self-supervised manner. On the next step, a semi-supervised classifier trained on the combination of labeled data from the first step and unlabeled data. The performance of the proposed framework will be compared with the results described earlier and also with optimized outcomes obtained from different optimization techniques.

Keywords: attention learning, language model, offensive language detection, self-supervised learning

Procedia PDF Downloads 106

Authors: Davut Disci

Abstract:

Objective: to measure the factors affecting expectations and intentions of using mobile phone in educational contexts by university students, using advanced equations and modeling techniques. Design and Methodology: According to the literature, Mobile Addiction, Parental Surveillance- Safety/Security, Social Relations, and Mobile Behavior are most used terms of defining mobile use of people. Therefore these variables are tried to be measured to find and estimate their effects on expectations and intentions of using mobile phone in educational context. 421 university students participated in this study and there are 229 Female and 192 Male students. For the purpose of examining the mobile behavior and educational expectations and intentions, a questionnaire is prepared and applied to the participants who had to answer all the questions online. Furthermore, responses to close-ended questions are analyzed by using The Statistical Package for Social Sciences(SPSS) software, reliabilities are measured by Cronbach’s Alpha analysis and hypothesis are examined via using Multiple Regression and Linear Regression analysis and the model is tested with Structural Equation Modeling(SEM) technique which is important for testing the model scientifically. Besides these responses, open-ended questions are taken into consideration. Results: When analyzing data gathered from close-ended questions, it is found that Mobile Addiction, Parental Surveillance, Social Relations and Frequency of Using Mobile Phone Applications are affecting the mobile behavior of the participants in different levels, helping them to use mobile phone in educational context. Moreover, as for open-ended questions, participants stated that they use many mobile applications in their learning environment in terms of contacting with friends, watching educational videos, finding course material via internet. They also agree in that mobile phone brings greater flexibility to their lives. According to the SEM results the model is not evaluated and it can be said that it may be improved to show in SEM besides in multiple regression. Conclusion: This study shows that the specified model can be used by educationalist, school authorities to improve their learning environment.

Keywords: education, mobile behavior, mobile learning, technology, Turkey

Procedia PDF Downloads 421

Authors: Khatuna Buskivadze

Abstract:

The aim of the present study is to investigate the sociolinguistic and classroom functions and frequency of Teacher’s Code Switching (CS) in the Content and Language Integrated (CLIL) Lesson. Nowadays, Georgian society struggles to become the part of the European world, the English language itself plays a role in forming new generations with European values. Based on our research conducted in 2019, out of all 114 private schools in Tbilisi, full- programs of CLIL are taught in 7 schools, while only some subjects using CLIL are conducted in 3 schools. The goal of the former research was to define the features of Content and Language Integrated learning (CLIL) methodology within the process of teaching English on the Example of Georgian private high schools. Taking the Georgian reality and cultural features into account, the modified version of the questionnaire, based on the classification of using CS in ESL Classroom proposed By Ferguson (2009) was used. The qualitative research revealed students’ and teacher’s attitudes towards teacher’s code-switching in CLIL lesson. Both qualitative and quantitative research were conducted: the observations of the teacher’s lessons (Recording of T’s online lessons), interview and the questionnaire among Math’s T’s 20 high school students. We came to the several conclusions, some of them are given here: Math’s teacher’s CS behavior mostly serves (1) the conversational function of interjection; (2) the classroom functions of introducing unfamiliar materials and topics, explaining difficult concepts, maintaining classroom discipline and the structure of the lesson; The teacher and 13 students have negative attitudes towards using only Georgian in teaching Math. The higher level of English is the more negative is attitude towards using Georgian in the classroom. Although all the students were Georgian, their competence in English is higher than in Georgian, therefore they consider English as an inseparable part of their identities. The overall results of the case study of teaching Math (Educational discourse) in one of the private schools in Tbilisi will be presented at the conference.

Keywords: attitudes, bilingualism, code-switching, CLIL, conversation analysis, interactional sociolinguistics.

Procedia PDF Downloads 162

Authors: Sihem Benaouda

Abstract:

English is gaining an international identity as it affects every academic and professional field in the world. Without increasing their cultural understanding, it would obviously be difficult to completely educate learners for communication in a globalised environment. The concept of culture is intricate and needs to be elucidated, especially in an English language teaching (ELT) context. The study focuses on the investigation of the cultural studies integrated into the different types of English for specific purposes (ESP) materials, as opposed to English for general purposes (EGP) textbooks. A qualitative methodology based on a triangulation of techniques was conducted through materials analysis of five textbooks in both advanced EGP and three types of ESP. In addition to a semi-structured interview conducted with Algerian ESP practitioners, data analysis results revealed that culture in ESP textbooks is not overtly isolated into chapters and that cultural studies are predominantly present in business and economics materials, namely English for hotel and catering staff, tourism, and flight attendants. However, implicit cultural instruction is signalled in the social sciences and is negligible in science and technology sources. In terms of content, cultural studies in EGP are more related to generic topics, whereas, in some ESP materials, the topics are rather oriented to the specific field they belong to. Furthermore, the respondents’ answers showed an unawareness of the importance of culture in ESP teaching, besides some disregard for culture teaching per se in ESP contexts.

Keywords: ESP, EGP, cultural studies, textbooks, teaching, materials

Procedia PDF Downloads 109

Authors: Sanskar, Abhinav Pal, Aryush Gupta, Sushil Kumar Mishra

Abstract:

One of the critical and tedious assignments in agricultural practices is the detection of diseases on vegetation. Agricultural production is very important in today’s economy because plant diseases are common, and early detection of plant diseases is important in agriculture. Automatic detection of such early diseases is useful because it reduces control efforts in large productive farms. Using digital image processing and machine learning algorithms, this paper presents a method for plant disease detection. Detection of the disease occurs on different leaves of the plant. The proposed system for plant disease detection is simple and computationally efficient, requiring less time than learning-based approaches. The accuracy of various plant and foliar diseases is calculated and presented in this paper.

Keywords: plant diseases, machine learning, image processing, deep learning

Procedia PDF Downloads 10

Authors: V. Nikkhah Rashidabad, M. Manteghian, M. Masoumi, S. Mousavian, D. Ashouri

Abstract:

In this research, the capability of neural networks in modeling and learning complicated and nonlinear relations has been used to develop a model for the prediction of changes in the diameter of bubbles in pool boiling distilled water. The input parameters used in the development of this network include element temperature, heat flux, and retention time of bubbles. The test data obtained from the experiment of the pool boiling of distilled water, and the measurement of the bubbles form on the cylindrical element. The model was developed based on training algorithm, which is typologically of back-propagation type. Considering the correlation coefficient obtained from this model is 0.9633. This shows that this model can be trusted for the simulation and modeling of the size of bubble and thermal transfer of boiling.

Keywords: bubble diameter, heat flux, neural network, training algorithm

Procedia PDF Downloads 443

Authors: Szymon Rutkowski, Tomasz Korbak

Abstract:

We present a reinforcement-learning-based method of training neural machine translation models without parallel corpora. The standard encoder-decoder approach to machine translation suffers from two problems we aim to address. First, it needs parallel corpora, which are scarce, especially for low-resource languages. Second, it lacks psychological plausibility of learning procedure: learning a foreign language is about learning to communicate useful information, not merely learning to transduce from one language’s 'encoding' to another. We instead pose the problem of learning to translate as learning a policy in a communication game between two agents: the translator and the classifier. The classifier is trained beforehand on a natural language inference task (determining the entailment relation between a premise and a hypothesis) in the target language. The translator produces a sequence of actions that correspond to generating translations of both the hypothesis and premise, which are then passed to the classifier. The translator is rewarded for classifier’s performance on determining entailment between sentences translated by the translator to disciple’s native language. Translator’s performance thus reflects its ability to communicate useful information to the classifier. In effect, we train a machine translation model without the need for parallel corpora altogether. While similar reinforcement learning formulations for zero-shot translation were proposed before, there is a number of improvements we introduce. While prior research aimed at grounding the translation task in the physical world by evaluating agents on an image captioning task, we found that using a linguistic task is more sample-efficient. Natural language inference (also known as recognizing textual entailment) captures semantic properties of sentence pairs that are poorly correlated with semantic similarity, thus enforcing basic understanding of the role played by compositionality. It has been shown that models trained recognizing textual entailment produce high-quality general-purpose sentence embeddings transferrable to other tasks. We use stanford natural language inference (SNLI) dataset as well as its analogous datasets for French (XNLI) and Polish (CDSCorpus). Textual entailment corpora can be obtained relatively easily for any language, which makes our approach more extensible to low-resource languages than traditional approaches based on parallel corpora. We evaluated a number of reinforcement learning algorithms (including policy gradients and actor-critic) to solve the problem of translator’s policy optimization and found that our attempts yield some promising improvements over previous approaches to reinforcement-learning based zero-shot machine translation.

Keywords: agent-based language learning, low-resource translation, natural language inference, neural machine translation, reinforcement learning

Procedia PDF Downloads 128

Authors: Aofan Liu, Qianqian Tan, Burra Venkata Durga Kumar

Abstract:

Data security is of the utmost importance in the healthcare area, as sensitive patient information is constantly sent around and analyzed by many different parties. The use of federated learning, which enables data to be evaluated locally on devices rather than being transferred to a central server, has emerged as a potential solution for protecting the privacy of user information. To protect against data breaches and unauthorized access, federated learning alone might not be adequate. In this context, the application of blockchain technology could provide the system extra protection. This study proposes a distributed federated learning system that is built on blockchain technology in order to enhance security in healthcare. This makes it possible for a wide variety of healthcare providers to work together on data analysis without raising concerns about the confidentiality of the data. The technical aspects of the system, including as the design and implementation of distributed learning algorithms, consensus mechanisms, and smart contracts, are also investigated as part of this process. The technique that was offered is a workable alternative that addresses concerns about the safety of healthcare while also fostering collaborative research and the interchange of data.

Keywords: data privacy, distributed system, federated learning, machine learning

Procedia PDF Downloads 134

Authors: Firas A. Al Laban, Mohamed Chabi, Sammani Danwawu Abdullahi

Abstract:

There are a real needs to integrate types of Open Educational Resources (OER) with an intelligent system to extract information and knowledge in the semantic searching level. Those needs raised because most of current learning standard adopted web based learning and the e-learning systems does not always serve all educational goals. Semantic Web systems provide educators, students, and researchers with intelligent queries based on a semantic knowledge management learning system. An ontology-based learning system is an advanced system, where ontology plays the core of the semantic web in a smart learning environment. The objective of this paper is to discuss the potentials of ontologies and mapping different kinds of ontologies; heterogeneous or homogenous to manage and control different types of Open Educational Resources. The important contribution of this research is to approach a methodology uses logical rules and conceptual relations to map between ontologies of different educational resources. We expect from this methodology to establish for an intelligent educational system supporting student tutoring, self and lifelong learning system.

Keywords: knowledge management systems, ontologies, semantic web, open educational resources

Procedia PDF Downloads 498

Authors: Yasser Aminifard, Hamdollah Askari

Abstract:

This study aimed at investigating any difference between Iranian senior high school students' performance on Academic Words (AWs) and Everyday-Life Words (ELWs). To this end, in the first phase, a number of 120 male senior high school students were randomly selected from among twelve high schools in Gachsaran to serve as the participants of the study. In the second phase, using purposive sampling, six high school teachers holding an MA in TEFL and with over twenty years of teaching experience were interviewed. Two multiple-choice tests, each comprising 40 items, were given to the participants in order to determine their performance on AWs and ELWs and follow-up semi-structured interviews were conducted to explore teachers' opinions about participants' performance on the two tests. To analyze the data, a paired-samples t-test was carried out to compare the results of both tests and the interviews were also transcribed to pinpoint important themes. The results of the t-test indicated that the participants performed significantly better on AWs than on ELWs. Additionally, results of the interviews boiled down to the fact that the English textbooks designed for Iranian high school students are fundamentally flawed on the grounds that there is a mismatch between students' real language learning needs and what is presented to them as "teaching-to-the-test" materials via these books. Finally, the implications and suggestions for further research are discussed.

Keywords: everyday-life words, academic words, textbooks, washback

Procedia PDF Downloads 456