Search results for: Text classification

Authors: Claudio Loconsole, Daniele Leonardis, Antonio Brunetti, Gianpaolo Francesco Trotta, Nicholas Caporusso, Vitoantonio Bevilacqua

Abstract:

Unfortunately, many people still do not have access to communication, with specific regard to reading and writing. Among them, people who are blind or visually impaired, have several difficulties in getting access to the world, compared to the sighted. Indeed, despite technology advancement and cost reduction, nowadays assistive devices are still expensive such as Braille-based input/output systems which enable reading and writing texts (e.g., personal notes, documents). As a consequence, assistive technology affordability is fundamental in supporting the visually impaired in communication, learning, and social inclusion. This, in turn, has serious consequences in terms of equal access to opportunities, freedom of expression, and actual and independent participation to a society designed for the sighted. Moreover, the visually impaired experience difficulties in recognizing objects and interacting with devices in any activities of daily living. It is not a case that Braille indications are commonly reported only on medicine boxes and elevator keypads. Several software applications for the automatic translation of written text into speech (e.g., Text-To-Speech - TTS) enable reading pieces of documents. However, apart from simple tasks, in many circumstances TTS software is not suitable for understanding very complicated pieces of text requiring to dwell more on specific portions (e.g., mathematical formulas or Greek text). In addition, the experience of reading\writing text is completely different both in terms of engagement, and from an educational perspective. Statistics on the employment rate of blind people show that learning to read and write provides the visually impaired with up to 80% more opportunities of finding a job. Especially in higher educational levels, where the ability to digest very complex text is key, accessibility and availability of Braille plays a fundamental role in reducing drop-out rate of the visually impaired, thus affecting the effectiveness of the constitutional right to get access to education. In this context, the Braille Lab project aims at overcoming these social needs by including affordability in designing and developing assistive tools for visually impaired people. In detail, our awarded project focuses on a technology innovation of the operation principle of existing assistive tools for the visually impaired leaving the Human-Machine Interface unchanged. This can result in a significant reduction of the production costs and consequently of tool selling prices, thus representing an important opportunity for social entrepreneurship. The first two assistive tools designed within the Braille Lab project following the proposed approach aims to provide the possibility to personally print documents and handouts and to read texts written in Braille using refreshable Braille display, respectively. The former, named ‘Braille Cartridge’, represents an alternative solution for printing in Braille and consists in the realization of an electronic-controlled dispenser printing (cartridge) which can be integrated within traditional ink-jet printers, in order to leverage the efficiency and cost of the device mechanical structure which are already being used. The latter, named ‘Braille Cursor’, is an innovative Braille display featuring a substantial technology innovation by means of a unique cursor virtualizing Braille cells, thus limiting the number of active pins needed for Braille characters.

Keywords: Human rights, social challenges and technology innovations, visually impaired, affordability, assistive tools

Procedia PDF Downloads 272

Authors: Joud Basyouni, Rama Zagzoog, Mashael Al Faleh, Jana Alireza

Abstract:

AI is changing many fields and speeding up tasks that used to take a long time. It used to take too long to edit photos. However, many AI-powered apps make photo editing, automatic effects, and animations much easier than other manual editing apps with no AI. The mobile app Photoleap edits photos and creates digital art using AI. Editing photos with text prompts is also becoming a standard these days with the help of apps like Photoleap. Now, users can change backgrounds, add animations, turn text into images, and create scenes with AI. This project report discusses the photo editing app's history and popularity. Photoleap resembles Photoshop, Canva, Photos, and Pixlr. The report includes survey questions to assess Photoleap user satisfaction. The report describes Photoleap's features and functions with screenshots. Photoleap uses AI well. Charts and graphs show Photoleap user ratings and comments from the survey. This project found that most Photoleap users liked how well it worked, was made, and was easy to use. People liked changing photos and adding backgrounds. Users can create stunning photo animations. A few users dislike the app's animations, AI art, and photo effects. The project report discusses the app's pros and cons and offers improvements.

Keywords: artificial intelligence, photoleap, images, background, photo editing

Procedia PDF Downloads 56

Authors: Lavanya Annadi, Fahimeh Jafari

Abstract:

Transportation brings immense benefits to society, but it also has its costs. Costs include such as the cost of infrastructure, personnel and equipment, but also the loss of life and property in traffic accidents on the road, delays in travel due to traffic congestion and various indirect costs in terms of air transport. More research has been done to identify the various factors that affect road accidents, such as road infrastructure, traffic, sociodemographic characteristics, land use, and the environment. The aim of this research is to predict the probabilistic crash prediction of vehicles using machine learning due to natural and structural reasons by excluding spontaneous reasons like overspeeding etc., in the United States. These factors range from weather factors, like weather conditions, precipitation, visibility, wind speed, wind direction, temperature, pressure, and humidity to human made structures like road structure factors like bump, roundabout, no exit, turning loop, give away, etc. Probabilities are dissected into ten different classes. All the predictions are based on multiclass classification techniques, which are supervised learning. This study considers all crashes that happened in all states collected by the US government. To calculate the probability, multinomial expected value was used and assigned a classification label as the crash probability. We applied three different classification models, including multiclass Logistic Regression, Random Forest and XGBoost. The numerical results show that XGBoost achieved a 75.2% accuracy rate which indicates the part that is being played by natural and structural reasons for the crash. The paper has provided in-deep insights through exploratory data analysis.

Keywords: road safety, crash prediction, exploratory analysis, machine learning

Procedia PDF Downloads 107

Authors: Heeyeul Kwon, Yongtae Park

Abstract:

Despite the highly touted benefits, emerging technologies have unleashed pervasive concerns regarding unintended and unforeseen social impacts. Thus, those wishing to create safe and socially acceptable products need to identify such side effects and mitigate them prior to the market proliferation. Various methodologies in the field of technology assessment (TA), namely Delphi, impact assessment, and scenario planning, have been widely incorporated in such a circumstance. However, literatures face a major limitation in terms of sole reliance on participatory workshop activities. They unfortunately missed out the availability of a massive untapped data source of futuristic information flooding through the Internet. This research thus seeks to gain insights into utilization of futuristic data, future-oriented documents from the Internet, as a supplementary method to generate social impact scenarios whilst capturing perspectives of experts from a wide variety of disciplines. To this end, network analysis is conducted based on the social keywords extracted from the futuristic documents by text mining, which is then used as a guide to produce a comprehensive set of detailed scenarios. Our proposed approach facilitates harmonized depictions of possible hazardous consequences of emerging technologies and thereby makes decision makers more aware of, and responsive to, broad qualitative uncertainties.

Keywords: emerging technologies, futuristic data, scenario, text mining

Procedia PDF Downloads 488

Authors: Yasmine Abu Adla, Racha Soubra, Milana Kasab, Mohamad O. Diab, Aly Chkeir

Abstract:

Over the past several years, researchers have shown a great interest in assessing the mobility of elderly people to measure their functional status. Usually, such an assessment is done by conducting tests that require the subject to walk a certain distance, turn around, and finally sit back down. Consequently, this study aims to provide an at home monitoring system to assess the patient’s status continuously. Thus, we proposed a technique to automatically detect when a subject sits down while walking at home. In this study, we utilized a Doppler radar system to capture the motion of the subjects. More than 20 features were extracted from the radar signals, out of which 11 were chosen based on their intraclass correlation coefficient (ICC > 0.75). Accordingly, the sequential floating forward selection wrapper was applied to further narrow down the final feature vector. Finally, 5 features were introduced to the linear discriminant analysis classifier, and an accuracy of 93.75% was achieved as well as a precision and recall of 95% and 90%, respectively.

Keywords: Doppler radar system, stand-to-sit phase, TUG test, machine learning, classification

Procedia PDF Downloads 157

Authors: Aadityaa Singla

Abstract:

The big five personality traits are as follows: openness, conscientiousness, extraversion, agreeableness, and neuroticism. In order to get an insight into their personality, many flocks to these categories, which each have different meanings/characteristics. This information is important not only to individuals but also to career professionals and psychologists who can use this information for candidate assessment or job recruitment. The links between AI and psychology have been well studied in cognitive science, but it is still a rather novel development. It is possible for various AI classification models to accurately predict a personality question via ten input questions. This would contrast with the hundred questions that normal humans have to answer to gain a complete picture of their five personality traits. In order to approach this problem, various AI classification models were used on a dataset to predict what a user may answer. From there, the model's prediction was compared to its actual response. Normally, there are five answer choices (a 20% chance of correct guess), and the models exceed that value to different degrees, proving their significance. By utilizing an MLP classifier, decision tree, linear model, and K-nearest neighbors, they were able to obtain a test accuracy of 86.643, 54.625, 47.875, and 52.125, respectively. These approaches display that there is potential in the future for more nuanced predictions to be made regarding personality.

Keywords: machine learning, personally, big five personality traits, cognitive science

Procedia PDF Downloads 143

Authors: Jessica Lopez Espejel, Mahaman Sanoussi Yahaya Alassan, Walid Dahhane, El Hassane Ettifouri

Abstract:

Pretrained transformer-based models have shown high performance in natural language generation tasks. However, a new wave of interest has surged: automatic programming language code generation. This task consists of translating natural language instructions to a source code. Despite the fact that well-known pre-trained models on language generation have achieved good performance in learning programming languages, effort is still needed in automatic code generation. In this paper, we introduce JaCoText, a model based on Transformer neural network. It aims to generate java source code from natural language text. JaCoText leverages the advantages of both natural language and code generation models. More specifically, we study some findings from state of the art and use them to (1) initialize our model from powerful pre-trained models, (2) explore additional pretraining on our java dataset, (3) lead experiments combining the unimodal and bimodal data in training, and (4) scale the input and output length during the fine-tuning of the model. Conducted experiments on CONCODE dataset show that JaCoText achieves new state-of-the-art results.

Keywords: java code generation, natural language processing, sequence-to-sequence models, transformer neural networks

Procedia PDF Downloads 276

Authors: Ali̇ Si̇nan Soğanci

Abstract:

The cement stabilization is one of the ground improvement method applied worldwide to increase the strength of clayey soils. The using of cement has got lots of advantages compared to other stabilization methods. Cement stabilization can be done quickly, the cost is low and creates a more durable structure with the soil. Cement can be used in the treatment of a wide variety of soils. The best results of the cement stabilization were seen on silts as well as coarse-grained soils. In this study, blocks of clay were taken from the Apa-Hotamış conveyance channel route which is 125km long will be built in Konya that take the water with 70m3/sec from Mavi tunnel to Hotamış storage. Firstly, the index properties of clay samples were determined according to the Unified Soil Classification System. The experimental program was carried out on compacted soil specimens with 0%, 7 %, 15% and 30 % cement additives and the results of unconfined compression strength were discussed. The results of unconfined compression tests indicated an increase in strength with increasing cement content.

Keywords: cement stabilization, unconfined compression test, clayey soils, unified soil classification system.

Procedia PDF Downloads 417

Authors: Abrar Ajmal

Abstract:

The language used in literature reveals the culture and social gestalt of any society in which it has been constructed and consumed. This paper carries the same rationale, which aims to track certain socio-religious and cultural-economic disparities and discrepancies towards minorities, particularly Christians, in an Islamic re(public) where there is a clear majority of Muslims with the help of analysis of instances of language used in the narratives “Our Lady of Alice Bhatt” by Mohammad Hanif. It would highlight social inequalities practiced deeply in sociocultural discourse. Moreover, this research would also touch upon the question of gender discrimination and gender construction as a female entity in a male-chauvinistic scenic turnout using language since the novel revolves around communicative forfeits of Alice Bhatti’s life where she is fraying in fisticuffs to befit herself in a miss-fitted society. It would employ using Fairclough's framework for analysis to conduct a critical discourse analysis of the text at three axiom levels namely textual analysis, discursive practices, and socio-cultural analysis. Thus, the results would reveal textual findings in linguistic analysis, a range of embedded discourses in discursive practices, and consumption of the text into socio-cultural explications with the use of language and lexicalization employed in the selected excerpts.

Keywords: gendered discourse, socio-economic disparities minorities, Islamization, analytical framework

Procedia PDF Downloads 53

Authors: Yu-Min Chang, Kuo-Sheng Tsai, Hung-Te Tsai, Chia-Hsin Li

Abstract:

This paper presents the early-warning lights classification management system for industrial parks promoted by the Taiwan Environmental Protection Administration (EPA) since 2011, including the definition of each early-warning light, objectives, action program and accomplishments. All of the 151 industrial parks in Taiwan were classified into four early-warning lights, including red, orange, yellow and green, for carrying out respective pollution management according to the monitoring data of soil and groundwater quality, regulatory compliance, and regulatory listing of control site or remediation site. The Taiwan EPA set up a priority list for high potential polluted industrial parks and investigated their soil and groundwater qualities based on the results of the light classification and pollution potential assessment. In 2011-2013, there were 44 industrial parks selected and carried out different investigation, such as the early warning groundwater well networks establishment and pollution investigation/verification for the red and orange-light industrial parks and the environmental background survey for the yellow-light industrial parks. Among them, 22 industrial parks were newly or continuously confirmed that the concentrations of pollutants exceeded those in soil or groundwater pollution control standards. Thus, the further investigation, groundwater use restriction, listing of pollution control site or remediation site, and pollutant isolation measures were implemented by the local environmental protection and industry competent authorities; the early warning lights of those industrial parks were proposed to adjust up to orange or red-light. Up to the present, the preliminary positive effect of the soil and groundwater quality management system for industrial parks has been noticed in several aspects, such as environmental background information collection, early warning of pollution risk, pollution investigation and control, information integration and application, and inter-agency collaboration. Finally, the work and goal of self-initiated quality management of industrial parks will be carried out on the basis of the inter-agency collaboration by the classified lights system of early warning and management as well as the regular announcement of the status of each industrial park.

Keywords: industrial park, soil and groundwater quality management, early-warning lights classification, SOP for reporting and treatment of monitored abnormal events

Procedia PDF Downloads 319

Authors: Leonardo Alfredo Mendoza, Cristian Munoz, Marco Aurelio Pacheco, Manoela Kohler, Evelyn Batista, Rodrigo Moura

Abstract:

Natural Language Processing (NLP) techniques are increasingly more powerful to be able to interpret the feelings and reactions of a person to a product or service. Sentiment analysis has become a fundamental tool for this interpretation but has few applications in languages other than English. This paper presents a classification of sentiment analysis in Portuguese with a base of comments from social networks in Portuguese. A word embedding's representation was used with a 50-Dimension GloVe pre-trained model, generated through a corpus completely in Portuguese. To generate this classification, the bidirectional long short-term memory and bidirectional Gated Recurrent Unit (GRU) models are used, reaching results of 99.1%.

Keywords: natural processing language, sentiment analysis, bidirectional long short-term memory, BI-LSTM, gated recurrent unit, GRU

Procedia PDF Downloads 154

Authors: Bowen Ding, Yihao Kuang

Abstract:

In recent years, large-scale pre-trained language models have achieved state-of-the-art performance on a variety of natural language processing tasks. The word vectors produced by these language models can be viewed as dense encoded presentations of natural language that in text form. However, it is unknown how much linguistic information is encoded and how. In this paper, we construct several corresponding probing tasks for multiple linguistic information to clarify the encoding capabilities of different language models and performed a visual display. We firstly obtain word presentations in vector form from different language models, including BERT, ELMo, RoBERTa and GPT. Classifiers with a small scale of parameters and unsupervised tasks are then applied on these word vectors to discriminate their capability to encode corresponding linguistic information. The constructed probe tasks contain both semantic and syntactic aspects. The semantic aspect includes the ability of the model to understand semantic entities such as numbers, time, and characters, and the grammatical aspect includes the ability of the language model to understand grammatical structures such as dependency relationships and reference relationships. We also compare encoding capabilities of different layers in the same language model to infer how linguistic information is encoded in the model.

Keywords: language models, probing task, text presentation, linguistic information

Procedia PDF Downloads 105

Authors: A. J. S. P. Nileema, S. Kulatunga , S. H. Palihawadana

Abstract:

Computer aided detection (CAD) system was developed for the diagnosis of pulmonary tuberculosis using digital chest X-rays with MATLAB image processing techniques using a statistical approach. The study comprised of 200 digital chest radiographs collected from the National Hospital for Respiratory Diseases - Welisara, Sri Lanka. Pre-processing was done to remove identification details. Lung fields were segmented and then divided into four quadrants; right upper quadrant, left upper quadrant, right lower quadrant, and left lower quadrant using the image processing techniques in MATLAB. Contrast, correlation, homogeneity, energy, entropy, and maximum probability texture features were extracted using the gray level co-occurrence matrix method. Descriptive statistics and normal distribution analysis were performed using SPSS. Depending on the radiologists’ interpretation, chest radiographs were classified manually into PTB - positive (PTBP) and PTB - negative (PTBN) classes. Features with standard normal distribution were analyzed using an independent sample T-test for PTBP and PTBN chest radiographs. Among the six features tested, contrast, correlation, energy, entropy, and maximum probability features showed a statistically significant difference between the two classes at 95% confidence interval; therefore, could be used in the classification of chest radiograph for PTB diagnosis. With the resulting value ranges of the five texture features with normal distribution, a classification algorithm was then defined to recognize and classify the quadrant images; if the texture feature values of the quadrant image being tested falls within the defined region, it will be identified as a PTBP – abnormal quadrant and will be labeled as ‘Abnormal’ in red color with its border being highlighted in red color whereas if the texture feature values of the quadrant image being tested falls outside of the defined value range, it will be identified as PTBN–normal and labeled as ‘Normal’ in blue color but there will be no changes to the image outline. The developed classification algorithm has shown a high sensitivity of 92% which makes it an efficient CAD system and with a modest specificity of 70%.

Keywords: chest radiographs, computer aided detection, image processing, pulmonary tuberculosis

Procedia PDF Downloads 122

Authors: Mohamed Mounir Tellache, Hiroyuki Kambara, Yasuharu Koike, Makoto Miyakoshi, Natsue Yoshimura

Abstract:

This study explores the practicality of using electroencephalographic (EEG) independent components to predict eight-direction finger movements in pseudo-real-time. Six healthy participants with individual-head MRI images performed finger movements in eight directions with two different arm configurations. The analysis was performed in two stages. The first stage consisted of using independent component analysis (ICA) to separate the signals representing brain activity from non-brain activity signals and to obtain the unmixing matrix. The resulting independent components (ICs) were checked, and those reflecting brain-activity were selected. Finally, the time series of the selected ICs were used to predict eight finger-movement directions using Sparse Logistic Regression (SLR). The second stage consisted of using the previously obtained unmixing matrix, the selected ICs, and the model obtained by applying SLR to classify a different EEG dataset. This method was applied to two different settings, namely the single-participant level and the group-level. For the single-participant level, the EEG dataset used in the first stage and the EEG dataset used in the second stage originated from the same participant. For the group-level, the EEG datasets used in the first stage were constructed by temporally concatenating each combination without repetition of the EEG datasets of five participants out of six, whereas the EEG dataset used in the second stage originated from the remaining participants. The average test classification results across datasets (mean ± S.D.) were 38.62 ± 8.36% for the single-participant, which was significantly higher than the chance level (12.50 ± 0.01%), and 27.26 ± 4.39% for the group-level which was also significantly higher than the chance level (12.49% ± 0.01%). The classification accuracy within [–45°, 45°] of the true direction is 70.03 ± 8.14% for single-participant and 62.63 ± 6.07% for group-level which may be promising for some real-life applications. Clustering and contribution analyses further revealed the brain regions involved in finger movement and the temporal aspect of their contribution to the classification. These results showed the possibility of using the ICA-based method in combination with other methods to build a real-time system to control prostheses.

Keywords: brain-computer interface, electroencephalography, finger motion decoding, independent component analysis, pseudo real-time motion decoding

Procedia PDF Downloads 137

Authors: L. S. Chathurika

Abstract:

Early prediction of student performance is an important factor to be gained academic excellence. Whatever the study stream in secondary education, students lay the foundation for higher studies during the first year of their degree or diploma program in Sri Lanka. The information technology (IT) field has certain improvements in the education domain by selecting specialization areas to show the talents and skills of students. These specializations can be software engineering, network administration, database administration, multimedia design, etc. After completing the first-year, students attempt to select the best path by considering numerous factors. The purpose of this experiment is to predict the best study path using machine learning algorithms. Five classification algorithms: decision tree, support vector machine, artificial neural network, Naïve Bayes, and logistic regression are selected and tested. The support vector machine obtained the highest accuracy, 82.4%. Then affecting features are recognized to select the best study path.

Keywords: algorithm, classification, evaluation, features, testing, training

Procedia PDF Downloads 118

Authors: G. Hafner

Abstract:

A method developed at the University of Stuttgart will be presented: ‘Analysis, Evaluation and Optimization of Food Management’. A major focus is represented by quantification of food losses and food waste as well as their classification and evaluation regarding a system optimization through waste prevention. For quantification and accounting of food, food losses and food waste along the food chain, a clear definition of core terms is required at the beginning. This includes their methodological classification and demarcation within sectors of the food value chain. The food chain is divided into agriculture, industry and crafts, trade and consumption (at home and out of home). For adjustment of core terms, the authors have cooperated with relevant stakeholders in Germany for achieving the goal of holistic and agreed definitions for the whole food chain. This includes modeling of sub systems within the food value chain, definition of terms, differentiation between food losses and food wastage as well as methodological approaches. ‘Food Losses’ and ‘Food Wastes’ are assigned to individual sectors of the food chain including a description of the respective methods. The method for analyzing, evaluation and optimization of food management systems consist of the following parts: Part I: Terms and Definitions. Part II: System Modeling. Part III: Procedure for Data Collection and Accounting Part. IV: Methodological Approaches for Classification and Evaluation of Results. Part V: Evaluation Parameters and Benchmarks. Part VI: Measures for Optimization. Part VII: Monitoring of Success The method will be demonstrated at the example of an invesigation of food losses and food wastage in the Federal State of Bavaria including an extrapolation of respective results to quantify food wastage in Germany.

Keywords: food losses, food waste, resource management, waste management, system analysis, waste minimization, resource efficiency

Procedia PDF Downloads 402

Authors: Mohammed Riyas Pp

Abstract:

This study aimed at investigating major issues in translating the Arabic Hadith terminologies into English, focusing on choosing the most appropriate translation for each, reviewing major Hadith works in English. This study is confined to twenty terminologies with regard to classification of Hadith based on authority, strength, number of transmitters and connections in Isnad. Almost all available translations are collected and analyzed to find the most proper translation based on linguistic and translational values. To the researcher, many translations lack precise understanding of either Hadith terminologies or English language and varieties of methodologies have influence on varieties of translations. This study provides a classification of translational and conceptual issues. Translational issues are related to translatability of these terminologies and their equivalence. Conceptual issues provide a list of misunderstandings due to wrong translations of terminologies. This study ends with a suggestion for unification in translating terminologies based on convention of Muslim scholars having good understanding of Hadith terminologies and English language.

Keywords: english language, hadith terminologies, equivalence in translation, problems in translation

Procedia PDF Downloads 184

Authors: Shahd Alshammari

Abstract:

This paper considers one of Margaret Atwood’s short stories ‘Lusus Naturae'. Through a critical lens that makes use of Julia Kristeva’s work on Powers of Horror and abjection, this paper suggests that the monstrous girl is the disabled woman, the abject in society. The monster is used as a metaphor for the unknown, the misunderstood, and the ‘different’ woman. Culturally Relevant Teaching (CRT) is a pedagogy that calls for making course material accessible and relevant to students. Through the study of literary texts, we are able to help create agency inside and outside the classroom. Stories are a necessary part of establishing connections across borders and boundaries. Stories are meant to raise awareness both inside and outside the classroom. The discussion is equally important, and the text is meant to facilitate relevant questions that the students need to consider when it comes to identity. Questions to consider are: what does it mean to be a ‘girl’ today, and what implications and consequences are at hand when you fail to perform this gendered identity? Gender is sometimes a fatal bond in the Middle East, and even more so, is the disability. In the case of our unnamed protagonist, she undergoes a process of un-becoming, a non-linear process of growing up. In a sense, it is a counter-Bildungsroman. The reading of this text emphasizes that a non-linear narrative is sometimes necessary for the female protagonist’s self-awareness and development. Discussion in class facilitates this sense of agency and questioning of gender and disability.

Keywords: disability, gender, literature, pedagogy

Procedia PDF Downloads 654

Authors: Oumaima Lahmar

Abstract:

This paper aims to define a structured topography for finance researchers seeking to navigate the body of knowledge in their extrapolation of finance phenomena. To make sense of the body of knowledge in finance, a probabilistic topic modeling approach is applied on 6000 abstracts of academic articles published in three top journals in finance between 1976 and 2020. This approach combines both machine learning techniques and natural language processing to statistically identify the conjunctions between research articles and their shared topics described each by relevant keywords. The topic modeling analysis reveals 35 coherent topics that can well depict finance literature and provide a comprehensive structure for the ongoing research themes. Comparing the extracted topics to the Journal of Economic Literature (JEL) classification system, a significant similarity was highlighted between the characterizing keywords. On the other hand, we identify other topics that do not match the JEL classification despite being relevant in the finance literature.

Keywords: finance literature, textual analysis, topic modeling, perplexity

Procedia PDF Downloads 166

Authors: Debarati Bhaumik, Diptish Dey

Abstract:

Multilevel models, increasingly deployed in industries such as insurance, food production, and entertainment within functions such as marketing and supply chain management, need to be transparent and ethical. Applications usually result in binary classification within groups or hierarchies based on a set of input features. Using open-source datasets, we demonstrate that popular explainability methods, such as SHAP and LIME, consistently underperform inaccuracy when interpreting these models. They fail to predict the order of feature importance, the magnitudes, and occasionally even the nature of the feature contribution (negative versus positive contribution to the outcome). Besides accuracy, the computational intractability of SHAP for binomial classification is a cause of concern. For transparent and ethical applications of these hierarchical statistical models, sound audit frameworks need to be developed. In this paper, we propose an audit framework for technical assessment of multilevel regression models focusing on three aspects: (i) model assumptions & statistical properties, (ii) model transparency using different explainability methods, and (iii) discrimination assessment. To this end, we undertake a quantitative approach and compare intrinsic model methods with SHAP and LIME. The framework comprises a shortlist of KPIs, such as PoCE (Percentage of Correct Explanations) and MDG (Mean Discriminatory Gap) per feature, for each of these three aspects. A traffic light risk assessment method is furthermore coupled to these KPIs. The audit framework will assist regulatory bodies in performing conformity assessments of AI systems using multilevel binomial classification models at businesses. It will also benefit businesses deploying multilevel models to be future-proof and aligned with the European Commission’s proposed Regulation on Artificial Intelligence.

Keywords: audit, multilevel model, model transparency, model explainability, discrimination, ethics

Procedia PDF Downloads 88

Authors: Christoph Linse, Thomas Martinetz

Abstract:

Recent findings have shown that Neural Networks generalize also in over-parametrized regimes with zero training error. This is surprising, since it is completely against traditional machine learning wisdom. In our empirical study we fortify these findings in the domain of fine-grained image classification. We show that very large Convolutional Neural Networks with millions of weights do learn with only a handful of training samples and without image augmentation, explicit regularization or pretraining. We train the architectures ResNet018, ResNet101 and VGG19 on subsets of the difficult benchmark datasets Caltech101, CUB_200_2011, FGVCAircraft, Flowers102 and StanfordCars with 100 classes and more, perform a comprehensive comparative study and draw implications for the practical application of CNNs. Finally, we show that VGG19 with 140 million weights learns to distinguish airplanes and motorbikes with up to 95% accuracy using only 20 training samples per class.

Keywords: convolutional neural networks, fine-grained image classification, generalization, image recognition, over-parameterized, small data sets

Procedia PDF Downloads 85

Authors: Wenbo Du, Xiaomei Ma

Abstract:

With the empowerment of Cognitive Diagnostic Assessment (CDA), various domains of language testing and assessment have been investigated to dig out more diagnostic information. What is noticeable is that most of the extant empirical CDA-based research puts much emphasis on individual-level diagnostic purpose with very few concerned about learners’ group-level performance. Even though the personalized diagnostic feedback is the unique feature that differentiates CDA from other assessment tools, group-level diagnostic information cannot be overlooked in that it might be more practical in classroom setting. Additionally, the group-level diagnostic information obtained via current CDA always results in a “flat pattern”, that is, the mastery/non-mastery of all tested skills accounts for the two highest proportion. In that case, the outcome does not bring too much benefits than the original total score. To address these issues, the present study attempts to apply cluster analysis for group classification and quantile regression analysis to pinpoint learners’ performance at different proficiency levels (beginner, intermediate and advanced) thus to enhance the interpretation of the CDA results extracted from a group of EFL learners’ reading performance on a diagnostic reading test designed by PELDiaG research team from a key university in China. The results show that EM method in cluster analysis yield more appropriate classification results than that of CDA, and quantile regression analysis does picture more insightful characteristics of learners with different reading proficiencies. The findings are helpful and practical for instructors to refine EFL reading curriculum and instructional plan tailored based on the group classification results and quantile regression analysis. Meanwhile, these innovative statistical methods could also make up the deficiencies of CDA and push forward the development of language testing and assessment in the future.

Keywords: cognitive diagnostic assessment, diagnostic feedback, EFL reading, quantile regression

Procedia PDF Downloads 145

Authors: Mohammad H. Taheri, David J. Brown, Nasser Sherkat

Abstract:

Engagement is one of the most important factors in determining successful outcomes and deep learning in students. Existing approaches to detect student engagement involve periodic human observations that are subject to inter-rater reliability. Our solution uses real-time multimodal multisensor data labeled by objective performance outcomes to infer the engagement of students. The study involves four students with a combined diagnosis of cerebral palsy and a learning disability who took part in a 3-month trial over 59 sessions. Multimodal multisensor data were collected while they participated in a continuous performance test. Eye gaze, electroencephalogram, body pose, and interaction data were used to create a model of student engagement through objective labeling from the continuous performance test outcomes. In order to achieve this, a type of continuous performance test is introduced, the Seek-X type. Nine features were extracted including high-level handpicked compound features. Using leave-one-out cross-validation, a series of different machine learning approaches were evaluated. Overall, the random forest classification approach achieved the best classification results. Using random forest, 93.3% classification for engagement and 42.9% accuracy for disengagement were achieved. We compared these results to outcomes from different models: AdaBoost, decision tree, k-Nearest Neighbor, naïve Bayes, neural network, and support vector machine. We showed that using a multisensor approach achieved higher accuracy than using features from any reduced set of sensors. We found that using high-level handpicked features can improve the classification accuracy in every sensor mode. Our approach is robust to both sensor fallout and occlusions. The single most important sensor feature to the classification of engagement and distraction was shown to be eye gaze. It has been shown that we can accurately predict the level of engagement of students with learning disabilities in a real-time approach that is not subject to inter-rater reliability, human observation or reliant on a single mode of sensor input. This will help teachers design interventions for a heterogeneous group of students, where teachers cannot possibly attend to each of their individual needs. Our approach can be used to identify those with the greatest learning challenges so that all students are supported to reach their full potential.

Keywords: affective computing in education, affect detection, continuous performance test, engagement, flow, HCI, interaction, learning disabilities, machine learning, multimodal, multisensor, physiological sensors, student engagement

Procedia PDF Downloads 86

Authors: Chad Goldsworthy, B. Rajeswari Matam

Abstract:

The use of deep learning for species identification in camera trap images has revolutionised our ability to study, conserve and monitor species in a highly efficient and unobtrusive manner, with state-of-the-art models achieving accuracies surpassing the accuracy of manual human classification. The high imbalance of camera trap datasets, however, results in poor accuracies for minority (rare or endangered) species due to their relative insignificance to the overall model accuracy. This paper investigates the use of Focal Loss, in comparison to the traditional Cross Entropy Loss function, to improve the identification of minority species in the “255 Bird Species” dataset from Kaggle. The results show that, although Focal Loss slightly decreased the accuracy of the majority species, it was able to increase the F1-score by 0.06 and improve the identification of the bottom two, five and ten (minority) species by 37.5%, 15.7% and 10.8%, respectively, as well as resulting in an improved overall accuracy of 2.96%.

Keywords: convolutional neural networks, data imbalance, deep learning, focal loss, species classification, wildlife conservation

Procedia PDF Downloads 185

Authors: Cíntia Martins Sanches

Abstract:

In the first part of Phoenissae of Seneca, Antigone is a guide to Oedipus, and they leave Thebes: he is blind searching for death (inflicting the punishment himself wished on the killer of Laius, ie exile and death); she is trying to convince him to give up such punishment and bring him back to Thebes. Concerning Oedipus lines, we observed a high frequency of Latin neuter in the treatment the protagonist gave to his daughter Antigone. We considered in this study that such frequency may be related to the sanctification of the daughter, who is seen by him as an enlightened being and without defects, free of the human condition (which takes on the existence of failures by essence). This study, thus, puts forward an analysis of the passages the said feature is present, relating them to the effect of meaning found in each occurrence. As part of a doctorate, this study investigates the stylistic idiom of Seneca in the Oedipus and Phoenissae tragedies, aiming at translating both tragedies expressively. The concept of stylistic idiom concerns the stylistic affinity required for a translation to be equivalent to the source text. In this wise, this study inquires into how the Latin text is organized poetically, pointing out the expressive features frequently appearing in both dramas. The method we used is based on the Semiotics theory — observing how connotation, ie a language use in which prevails the poetic function, naturally polysemous, acts to achieve each expressive effect.

Keywords: antigone, neuter, Oedipus, Phoenissae, Seneca

Procedia PDF Downloads 284

Authors: Sihem Oujdi, Hafida Belbachir

Abstract:

Existing methods of data mining cannot be applied on spatial data because they require spatial specificity consideration, as spatial relationships. This paper focuses on the classification with decision trees, which are one of the data mining techniques. We propose an extension of the C4.5 algorithm for spatial data, based on two different approaches Join materialization and Querying on the fly the different tables. Similar works have been done on these two main approaches, the first - Join materialization - favors the processing time in spite of memory space, whereas the second - Querying on the fly different tables- promotes memory space despite of the processing time. The modified C4.5 algorithm requires three entries tables: a target table, a neighbor table, and a spatial index join that contains the possible spatial relationship among the objects in the target table and those in the neighbor table. Thus, the proposed algorithms are applied to a spatial data pattern in the accidentology domain. A comparative study of our approach with other works of classification by spatial decision trees will be detailed.

Keywords: C4.5 algorithm, decision trees, S-CART, spatial data mining

Procedia PDF Downloads 610

Authors: Pedro G. Rodrigues, Maria J. Bárrios, Sara A. Ambrósio

Abstract:

The digitalisation of health is a profoundly transformative economic, political, and social process. As is often the case, such processes need to be carefully managed if misunderstandings, policy misalignments, or outright conflicts between the government and a wide gamut of stakeholders with competing interests are to be avoided. Thus, ensuring open lines of communication where all parties know what each other’s concerns are is key to good governance, as well as efficient and effective policymaking. This project aims to make a small but still significant contribution in this regard in that we seek to determine the extent to which health managers’ perceptions of what is a priority for digital health in Portugal are aligned with official policy perspectives. By applying state-of-the-art artificial intelligence technology first to the indexed literature on digital health and then to a set of official policy documents on the same topic, followed by a survey directed at health managers working in public and private hospitals in Portugal, we obtain two priority rankings that, when compared, will allow us to produce a synthesis and toolkit on digital health policy in Portugal, with a view to identifying areas of policy convergence and divergence. This project is also particularly peculiar in the sense that sophisticated digital methods related to text analytics are employed to study good governance aspects of digitalisation applied to health care.

Keywords: digital health, health informatics, text analytics, governance, natural language understanding

Procedia PDF Downloads 60

Authors: R. Periyasamy, Deepak Joshi, Sneh Anand

Abstract:

Foot posture assessment is important to evaluate foot type, causing gait and postural defects in all age groups. Although different methods are used for classification of foot arch type in clinical/research examination, there is no clear approach for selecting the most appropriate measurement system. Therefore, the aim of this study was to develop a system for evaluation of foot type as clinical decision-making aids for diagnosis of flat and normal arch based on the Arch Index (AI) and foot pressure distribution parameter - Power Ratio (PR) data. The accuracy of the system was evaluated for 27 subjects with age ranging from 24 to 65 years. Foot area measurements (hind foot, mid foot, and forefoot) were acquired simultaneously from foot pressure intensity image using portable PedoPowerGraph system and analysis of the image in frequency domain to obtain foot pressure distribution parameter - PR data. From our results, we obtain 100% classification accuracy of normal and flat foot by using the linear discriminant analysis method. We observe there is no misclassification of foot types because of incorporating foot pressure distribution data instead of only arch index (AI). We found that the mid-foot pressure distribution ratio data and arch index (AI) value are well correlated to foot arch type based on visual analysis. Therefore, this paper suggests that the proposed system is accurate and easy to determine foot arch type from arch index (AI), as well as incorporating mid-foot pressure distribution ratio data instead of physical area of contact. Hence, such computational tool based system can help the clinicians for assessment of foot structure and cross-check their diagnosis of flat foot from mid-foot pressure distribution.

Keywords: arch index, computational tool, static foot pressure intensity image, foot pressure distribution, linear discriminant analysis

Procedia PDF Downloads 494

Authors: Kefaya Qaddoum

Abstract:

Most of greenhouse growers desire a determined amount of yields in order to accurately meet market requirements. The purpose of this paper is to model a simple but often satisfactory supervised classification method. The original naive Bayes have a serious weakness, which is producing redundant predictors. In this paper, utilized regularization technique was used to obtain a computationally efficient classifier based on naive Bayes. The suggested construction, utilized L1-penalty, is capable of clearing redundant predictors, where a modification of the LARS algorithm is devised to solve this problem, making this method applicable to a wide range of data. In the experimental section, a study conducted to examine the effect of redundant and irrelevant predictors, and test the method on WSG data set for tomato yields, where there are many more predictors than data, and the urge need to predict weekly yield is the goal of this approach. Finally, the modified approach is compared with several naive Bayes variants and other classification algorithms (SVM and kNN), and is shown to be fairly good.

Keywords: tomato yield prediction, naive Bayes, redundancy, WSG

Procedia PDF Downloads 228

Authors: H. J. Wattimanela, U. S. Passaribu, A. N. T. Puspito, S. W. Indratno

Abstract:

Molluca Collision Zone is located at the junction of the Eurasian plate, Australian, Pacific, and the Philippines. Between the Sangihe arc, west of the collision zone, and to the east of Halmahera arc is active collision and convex toward the Molluca Sea. This research will analyze the behavior of earthquake occurrence in Molluca Collision Zone related to the distributions of an earthquake in each partition regions, determining the type of distribution of a occurrence earthquake of partition regions, and the mean occurrence of earthquakes each partition regions, and the correlation between the partitions region. We calculate number of earthquakes using partition method and its behavioral using conventional statistical methods. The data used is the data type of shallow earthquakes with magnitudes ≥ 4 SR for the period 1964-2013 in the Molluca Collision Zone. From the results, we can classify partitioned regions based on the correlation into two classes: strong and very strong. This classification can be used for early warning system in disaster management.

Keywords: molluca collision zone, partition regions, conventional statistical methods, earthquakes, classifications, disaster management

Procedia PDF Downloads 490