Search results for: Associative Classification

Authors: Tahani Aljohani, Jialin Yu, Alexandra. I. Cristea

Abstract:

The more an educational system knows about a learner, the more personalised interaction it can provide, which leads to better learning. However, asking a learner directly is potentially disruptive, and often ignored by learners. Especially in the booming realm of MOOC Massive Online Learning platforms, only a very low percentage of users disclose demographic information about themselves. Thus, in this paper, we aim to predict learners’ demographic characteristics, by proposing an approach using linguistically motivated Deep Learning Architectures for Learner Profiling, particularly targeting gender prediction on a FutureLearn MOOC platform. Additionally, we tackle here the difficult problem of predicting the gender of learners based on their comments only – which are often available across MOOCs. The most common current approaches to text classification use the Long Short-Term Memory (LSTM) model, considering sentences as sequences. However, human language also has structures. In this research, rather than considering sentences as plain sequences, we hypothesise that higher semantic - and syntactic level sentence processing based on linguistics will render a richer representation. We thus evaluate, the traditional LSTM versus other bleeding edge models, which take into account syntactic structure, such as tree-structured LSTM, Stack-augmented Parser-Interpreter Neural Network (SPINN) and the Structure-Aware Tag Augmented model (SATA). Additionally, we explore using different word-level encoding functions. We have implemented these methods on Our MOOC dataset, which is the most performant one comparing with a public dataset on sentiment analysis that is further used as a cross-examining for the models' results.

Keywords: Deep learning, data mining, gender predication, MOOCs.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1363

Authors: K. Nirmala Devi, V. Murali Bhaskaran

Abstract:

Traditional document representation for classification follows Bag of Words (BoW) approach to represent the term weights. The conventional method uses the Vector Space Model (VSM) to exploit the statistical information of terms in the documents and they fail to address the semantic information as well as order of the terms present in the documents. Although, the phrase based approach follows the order of the terms present in the documents rather than semantics behind the word. Therefore, a semantic concept based approach is used in this paper for enhancing the semantics by incorporating the ontology information. In this paper a novel method is proposed to forecast the intraday stock market price directional movement based on the sentiments from Twitter and money control news articles. The stock market forecasting is a very difficult and highly complicated task because it is affected by many factors such as economic conditions, political events and investor’s sentiment etc. The stock market series are generally dynamic, nonparametric, noisy and chaotic by nature. The sentiment analysis along with wisdom of crowds can automatically compute the collective intelligence of future performance in many areas like stock market, box office sales and election outcomes. The proposed method utilizes collective sentiments for stock market to predict the stock price directional movements. The collective sentiments in the above social media have powerful prediction on the stock price directional movements as up/down by using Granger Causality test.

Keywords: Bag of Words, Collective Sentiments, Ontology, Semantic relations, Sentiments, Social media, Stock Prediction, Twitter, Vector Space Model and wisdom of crowds.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2801

Authors: Gholamreza Habibi

Abstract:

In order to study seed yield and seed yield components in bean under reduced irrigation condition and assessment drought tolerance of genotypes, 15 lines of White beans were evaluated in two separate RCB design with 3 replications under stress and non stress conditions. Analysis of variance showed that there were significant differences among varieties in terms of traits under study, indicating the existence of genetic variation among varieties. The results indicate that drought stress reduced seed yield, number of seed per plant, biological yield and number of pod in White been. In non stress condition, yield was highly correlated with the biological yield, whereas in stress condition it was highly correlated with harvest index. Results of stepwise regression showed that, selection can we done based on, biological yield, harvest index, number of seed per pod, seed length, 100 seed weight. Result of path analysis showed that the highest direct effect, being positive, was related to biological yield in non stress and to harvest index in stress conditions. Factor analysis were accomplished in stress and nonstress condition a, there were 4 factors that explained more than 76 percent of total variations. We used several selection indices such as Stress Susceptibility Index ( SSI ), Geometric Mean Productivity ( GMP ), Mean Productivity ( MP ), Stress Tolerance Index ( STI ) and Tolerance Index ( TOL ) to study drought tolerance of genotypes, we found that the best Stress Index for selection tolerance genotypes were STI, GMP and MP were the greatest correlations between these Indices and seed yield under stress and non stress conditions. In classification of genotypes base on phenotypic characteristics, using cluster analysis ( UPGMA ), all allels classified in 5 separate groups in stress and non stress conditions.

Keywords: Cluster analysis, factor analysis, path analysis, selection index, White bean

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2140

Authors: Siripen Yiamjanya, Kevin Wongleedee

Abstract:

This research paper aims to identify travel motivation by push and pull factors that affected decision making of international tourists in selecting Thailand as their destination choice. A total of 200 international tourists who traveled to Thailand during January and February, 2014 were used as the sample in this study. A questionnaire was employed as a tool in collecting the data, conducted in Bangkok. The list consisted of 30 attributes representing both psychological factors as “push- based factors” and destination factors as “pull-based factors”. Mean and standard deviation were used in order to find the top ten travel motives that were important determinants in the respondents’ decision making process to select Thailand as their destination choice. The finding revealed the top ten travel motivations influencing international tourists to select Thailand as their destination choice included [i] getting experience in foreign land; [ii] Thai food; [iii] learning new culture; [iv] relaxing in foreign land; [v] wanting to learn new things; [vi] being interested in Thai culture, and traditional markets; [vii] escaping from same daily life; [viii] enjoying activities; [ix] adventure; and [x] good weather. Classification of push- based and pull- based motives suggested that getting experience in foreign land was the most important push motive for international tourists to travel, while Thai food portrayed its highest significance as pull motive. Discussion and suggestions were also made for tourism industry of Thailand.

Keywords: Decision Making, Destination Choice, International Tourist, Pull Factor, Push Factor, Thailand, Travel Motivation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 16386

Authors: S. Abdi Goudazri, R. Ziaie Moayed, A. Nazeri

Abstract:

Loose soils normally are of weak bearing capacity due to their structural nature. Being exposed to heavy traffic loads, they would fail in most cases. To tackle the aforementioned issue, geotechnical engineers have come up with different approaches; one of which is making use of geosynthetic-reinforced soil-aggregate systems. As these polymeric reinforcements have highlighted economic and environmentally-friendly features, they have become widespread in practice during the last decades. The present research investigates the efficiency of four different types of these reinforcements in increasing the bearing capacity of two-layered soil sections using a series California Bearing Ratio (CBR) test. The studied sections are comprised of a 10 cm-thick layer of no. 161 Firouzkooh sand (weak subgrade) and a 10 cm-thick layer of compacted aggregate materials (base course) classified as SP and GW according to the United Soil Classification System (USCS), respectively. The aggregate layer was compacted to the relative density (Dr) of 95% at the optimum water content (Wopt) of 6.5%. The applied reinforcements were including two kinds of geocomposites (type A and B), a geotextile, and a geogrid that were embedded at the interface of the lower and the upper layers of the soil-aggregate system. As the standard CBR mold was not appropriate in height for this study, the mold used for soaked CBR tests were utilized. To make a comparison between the results of stress-settlement behavior in the studied specimens, CBR values pertinent to the penetrations of 2.5 mm and 5 mm were considered. The obtained results demonstrated 21% and 24.5% increments in the amount of CBR value in the presence of geocomposite type A and geogrid, respectively. On the other hand, the effect of both geotextile and geocomposite type B on CBR values was generally insignificant in this research.

Keywords: Geosynthetics, geogrid, geotextile, CBR test, increasing bearing capacity.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 692

Authors: O. Onoriode Emoyan, E. Eyitemi Akporhonor, Charles Otobrise

Abstract:

Polycyclic Aromatic Hydrocarbons (PAHs) are formed mainly because of incomplete combustion of organic materials during industrial, domestic activities or natural occurrence. Their toxicity and contamination of terrestrial and aquatic ecosystem have been established. However, with limited validity index, previous research has focused on PAHs isomer pair ratios of variable physicochemical properties in source identification. The objective of this investigation was to determine the empirical validity of Pearson Correlation Coefficient (PCC) and Cluster Analysis (CA) in PAHs source identification along soil samples of different land uses. Therefore, 16 PAHs grouped, as Endocrine Disruption Substances (EDSs) were determined in 10 sample stations in top and sub soils seasonally. PAHs was determined the use of Varian 300 gas chromatograph interfaced with flame ionization detector. Instruments and reagents used are of standard and chromatographic grades respectively. PCC and CA results showed that the classification of PAHs along pyrolitic and petrogenic organics used in source signature is about the predominance PAHs in environmental matrix. Therefore, the distribution of PAHs in the studied stations revealed the presence of trace quantities of the vast majority of the sixteen PAHs, which may ultimately inhabit the actual source signature authentication. Therefore, factors to be considered when evaluating possible sources of PAHs could be; type and extent of bacterial metabolism, transformation products/substrates, and environmental factors such as salinity, pH, oxygen concentration, nutrients, light intensity, temperature, co-substrates, and environmental medium are hereby recommended as factors to be considered when evaluating possible sources of PAHs.

Keywords: Comparative correlation, kinetically, polynuclear aromatic hydrocarbons, thermodynamically- favored PAHs, sources evaluation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1984

Authors: Cheng Zeng, George Michailidis, Hitoshi Iyatomi, Leo L Duan

Abstract:

The conditional density characterizes the distribution of a response variable y given other predictor x, and plays a key role in many statistical tasks, including classification and outlier detection. Although there has been abundant work on the problem of Conditional Density Estimation (CDE) for a low-dimensional response in the presence of a high-dimensional predictor, little work has been done for a high-dimensional response such as images. The promising performance of normalizing flow (NF) neural networks in unconditional density estimation acts a motivating starting point. In this work, we extend NF neural networks when external x is present. Specifically, they use the NF to parameterize a one-to-one transform between a high-dimensional y and a latent z that comprises two components [zP , zN]. The zP component is a low-dimensional subvector obtained from the posterior distribution of an elementary predictive model for x, such as logistic/linear regression. The zN component is a high-dimensional independent Gaussian vector, which explains the variations in y not or less related to x. Unlike existing CDE methods, the proposed approach, coined Augmented Posterior CDE (AP-CDE), only requires a simple modification on the common normalizing flow framework, while significantly improving the interpretation of the latent component, since zP represents a supervised dimension reduction. In image analytics applications, AP-CDE shows good separation of x-related variations due to factors such as lighting condition and subject id, from the other random variations. Further, the experiments show that an unconditional NF neural network, based on an unsupervised model of z, such as Gaussian mixture, fails to generate interpretable results.

Keywords: Conditional density estimation, image generation, normalizing flow, supervised dimension reduction.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 167

Authors: Isaac K. E. Ampomah, Seong-Bae Park, Sang-Jo Lee

Abstract:

Over the past decade, there have been promising developments in Natural Language Processing (NLP) with several investigations of approaches focusing on Recognizing Textual Entailment (RTE). These models include models based on lexical similarities, models based on formal reasoning, and most recently deep neural models. In this paper, we present a sentence encoding model that exploits the sentence-to-sentence relation information for RTE. In terms of sentence modeling, Convolutional neural network (CNN) and recurrent neural networks (RNNs) adopt different approaches. RNNs are known to be well suited for sequence modeling, whilst CNN is suited for the extraction of n-gram features through the filters and can learn ranges of relations via the pooling mechanism. We combine the strength of RNN and CNN as stated above to present a unified model for the RTE task. Our model basically combines relation vectors computed from the phrasal representation of each sentence and final encoded sentence representations. Firstly, we pass each sentence through a convolutional layer to extract a sequence of higher-level phrase representation for each sentence from which the first relation vector is computed. Secondly, the phrasal representation of each sentence from the convolutional layer is fed into a Bidirectional Long Short Term Memory (Bi-LSTM) to obtain the final sentence representations from which a second relation vector is computed. The relations vectors are combined and then used in then used in the same fashion as attention mechanism over the Bi-LSTM outputs to yield the final sentence representations for the classification. Experiment on the Stanford Natural Language Inference (SNLI) corpus suggests that this is a promising technique for RTE.

Keywords: Deep neural models, natural language inference, recognizing textual entailment, sentence-to-sentence relation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1454

Authors: P. Massanyi, L. Kichi, T. Slanina, E. Kolesar, J. Danko, N. Lukac, E. Tvrda, R. Stawarz, A. Kolesarova

Abstract:

Target of this study was the analysis of the impact of crude glycerol on canine spermatozoa motility, morphology, viability, and membrane integrity. Experiments were realized in vitro. In the study, semen from 5 large dog breeds was used. They were typical representatives of large breeds, coming from healthy rearing, regularly vaccinated and integrated to the further breeding. Semen collections were realized at the owners of animals and in the veterinary clinic. Subsequently the experiments were realized at the Department of Animal Physiology of the SUA in Nitra. The spermatozoa motility was evaluated using CASA analyzer (SpermVisionTM, Minitub, Germany) at the temperature 5 and 37°C for 5 hours. In the study, 13 motility parameters were evaluated. Generally, crude glycerol has generally negative effect on spermatozoa motility. Morphological analysis was realized using Hancock staining and the preparations were evaluated at magnification 1000x using classification tables of morphologically changed spermatozoa. Data clearly detected the highest number of morphologically changed spermatozoa in the experimental groups (know twisted tails, tail torso and tail coiling). For acrosome alterations swelled acrosomes, removed acrosomes and acrosomes with undulated membrane were detected. In this study also the effect of crude glycerol on spermatozoa membrane integrity were analyzed. The highest crude glycerol concentration significantly affects spermatozoa integrity. Results of this study show that crude glycerol has effect of spermatozoa motility, viability, and membrane integrity. Detected changes are related to crude glycerol concentration, temperature, as well as time of incubation.

Keywords: Dog, semen, spermatozoa, acrosome, glycerol, CASA, viability.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2102

Authors: Joe Walsh, Mike Climstein, Ian Timothy Heazlewood, Stephen Burke, Jyrki Kettunen, Kent Adams, Mark DeBeliso

Abstract:

Thousands of masters athletes participate quadrennially in the World Masters Games (WMG), yet this cohort of athletes remains proportionately under-investigated. Due to a growing global obesity pandemic in context of benefits of physical activity across the lifespan, the BMI trends for this unique population was of particular interest. The nexus between health, physical activity and aging is complex and has raised much interest in recent times due to the realization that a multifaceted approach is necessary in order to counteract the obesity pandemic. By investigating age based trends within a population adhering to competitive sport at older ages, further insight might be gleaned to assist in understanding one of many factors influencing this relationship.BMI was derived using data gathered on a total of 6,071 masters athletes (51.9% male, 48.1% female) aged 25 to 91 years ( =51.5, s =±9.7), competing at the Sydney World Masters Games (2009). Using linear and loess regression it was demonstrated that the usual tendency for prevalence of higher BMI increasing with age was reversed in the sample. This trend in reversal was repeated for both male and female only sub-sets of the sample participants, indicating the possibility of improved prevalence of BMI with increasing age for both the sample as a whole and these individual sub-groups.This evidence of improved classification in one index of health (reduced BMI) for masters athletes (when compared to the general population) implies there are either improved levels of this index of health with aging due to adherence to sport or possibly the reduced BMI is advantageous and contributes to this cohort adhering (or being attracted) to masters sport at older ages.

Keywords: Aging, masters athlete, Quetelet Index, sport

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1712

Authors: Javier Romera, Alberto Justo, Ignacio Fidalgo, Javier Araluce, Joshué Pérez

Abstract:

Nowadays, individuals with mobility needs face a significant challenge when docking vehicles. In many cases, after parking, they encounter insufficient space to exit, leading to two undesired outcomes: either avoiding parking in that spot or settling for improperly placed vehicles. To address this issue, this paper presents a parking control system employing gestural teleoperation. The system comprises three main phases: capturing body markers, interpreting gestures, and transmitting orders to the vehicle. The initial phase is centered around the MediaPipe framework, a versatile tool optimized for real-time gesture recognition. MediaPipe excels at detecting and tracing body markers, with a special emphasis on hand gestures. Hands detection is done by generating 21 reference points for each hand. Subsequently, after data capture, the project employs the MultiPerceptron Layer (MPL) for in-depth gesture classification. This tandem of MediaPipe’s extraction prowess and MPL’s analytical capability ensures that human gestures are translated into actionable commands with high precision. Furthermore, the system has been trained and validated within a built-in dataset. To prove the domain adaptation, a framework based on the Robot Operating System 2 (ROS2), as a communication backbone, alongside CARLA Simulator, is used. Following successful simulations, the system is transitioned to a real-world platform, marking a significant milestone in the project. This real-vehicle implementation verifies the practicality and efficiency of the system beyond theoretical constructs.

Keywords: Gesture detection, MediaPipe, MultiLayer Perceptron Layer, Robot Operating System.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 140

Authors: Svetlana Cojocaru, Mircea Petic, Inga Titchiev

Abstract:

Humanity faces more and more often with different social disasters, which in turn can generate new accidents and catastrophes. To mitigate their consequences, it is important to obtain early possible signals about the events which are or can occur and to prepare the corresponding scenarios that could be applied. Our research is focused on solving two problems in this domain: identifying signals related that an accident occurred or may occur and mitigation of some consequences of disasters. To solve the first problem, methods of selecting and processing texts from global network Internet are developed. Information in Romanian is of special interest for us. In order to obtain the mentioned tools, we should follow several steps, divided into preparatory stage and processing stage. Throughout the first stage, we manually collected over 724 news articles and classified them into 10 categories of social disasters. It constitutes more than 150 thousand words. Using this information, a controlled vocabulary of more than 300 keywords was elaborated, that will help in the process of classification and identification of the texts related to the field of social disasters. To solve the second problem, the formalism of Petri net has been used. We deal with the problem of inhabitants’ evacuation in useful time. The analysis methods such as reachability or coverability tree and invariants technique to determine dynamic properties of the modeled systems will be used. To perform a case study of properties of extended evacuation system by adding time, the analysis modules of PIPE such as Generalized Stochastic Petri Nets (GSPN) Analysis, Simulation, State Space Analysis, and Invariant Analysis have been used. These modules helped us to obtain the average number of persons situated in the rooms and the other quantitative properties and characteristics related to its dynamics.

Keywords: Lexicon of disasters, modelling, Petri nets, text annotation, social disasters.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1158

Authors: V. Veeraprathap, G. S. Harish, G. Narendra Kumar

Abstract:

Uncontrolled growth of abnormal cells in the lung in the form of tumor can be either benign (non-cancerous) or malignant (cancerous). Patients with Lung Cancer (LC) have an average of five years life span expectancy provided diagnosis, detection and prediction, which reduces many treatment options to risk of invasive surgery increasing survival rate. Computed Tomography (CT), Positron Emission Tomography (PET), and Magnetic Resonance Imaging (MRI) for earlier detection of cancer are common. Gaussian filter along with median filter used for smoothing and noise removal, Histogram Equalization (HE) for image enhancement gives the best results without inviting further opinions. Lung cavities are extracted and the background portion other than two lung cavities is completely removed with right and left lungs segmented separately. Region properties measurements area, perimeter, diameter, centroid and eccentricity measured for the tumor segmented image, while texture is characterized by Gray-Level Co-occurrence Matrix (GLCM) functions, feature extraction provides Region of Interest (ROI) given as input to classifier. Two levels of classifications, K-Nearest Neighbor (KNN) is used for determining patient condition as normal or abnormal, while Artificial Neural Networks (ANN) is used for identifying the cancer stage is employed. Discrete Wavelet Transform (DWT) algorithm is used for the main feature extraction leading to best efficiency. The developed technology finds encouraging results for real time information and on line detection for future research.

Keywords: ANN, DWT, GLCM, KNN, ROI, artificial neural networks, discrete wavelet transform, gray-level co-occurrence matrix, k-nearest neighbor, region of interest.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 960

Authors: Walsh Joe, Climstein Mike, Heazlewood Ian Timothy, Burke Stephen, Kettunen Jyrki, Adams Kent, DeBeliso Mark

Abstract:

Whilst there is growing evidence that activity across the lifespan is beneficial for improved health, there are also many changes involved with the aging process and subsequently the potential for reduced indices of health. The nexus between health, physical activity and aging is complex and has raised much interest in recent times due to the realization that a multifaceted approached is necessary in order to counteract a growing obesity epidemic. By investigating age based trends within a population adhering to competitive sport at older ages, further insight might be gleaned to assist in understanding one of many factors influencing this relationship. BMI was derived using data gathered on a total of 6,071 masters athletes (51.9% male, 48.1% female) aged 25 to 91 years ( =51.5, s =±9.7), competing at the Sydney World Masters Games (2009). Using linear and loess regression it was demonstrated that the usual tendency for prevalence of higher BMI increasing with age was reversed in the sample. This trend in reversal was repeated for both male and female only sub-sets of the sample participants, indicating the possibility of improved prevalence of BMI with increasing age for both the sample as a whole and these individual subgroups. This evidence of improved classification in one index of health (reduced BMI) for masters athletes (when compared to the general population) implies there are either improved levels of this index of health with aging due to adherence to sport or possibly the reduced BMI is advantageous and contributes to this cohort adhering (or being attracted) to masters sport at older ages. Demonstration of this proportionately under-investigated World Masters Games population having an improved relationship between BMI and increasing age over the general population is of particular interest in the context of the measures being taken globally to curb an obesity epidemic.

Keywords: Aging, masters athlete, Quetelet Index, sport.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1677

Authors: E. Feizi Ashtiani, S. Jamshidi, M.H Niksokhan, A. Feizi Ashtiani

Abstract:

Waste load allocation (WLA) policies may use multiobjective optimization methods to find the most appropriate and sustainable solutions. These usually intend to simultaneously minimize two criteria, total abatement costs (TC) and environmental violations (EV). If other criteria, such as inequity, need for minimization as well, it requires introducing more binary optimizations through different scenarios. In order to reduce the calculation steps, this study presents value index as an innovative decision making approach. Since the value index contains both the environmental violation and treatment costs, it can be maximized simultaneously with the equity index. It implies that the definition of different scenarios for environmental violations is no longer required. Furthermore, the solution is not necessarily the point with minimized total costs or environmental violations. This idea is testified for Haraz River, in north of Iran. Here, the dissolved oxygen (DO) level of river is simulated by Streeter-Phelps equation in MATLAB software. The WLA is determined for fish farms using multi-objective particle swarm optimization (MOPSO) in two scenarios. At first, the trade-off curves of TC-EV and TC-Inequity are plotted separately as the conventional approach. In the second, the Value-Equity curve is derived. The comparative results show that the solutions are in a similar range of inequity with lower total costs. This is due to the freedom of environmental violation attained in value index. As a result, the conventional approach can well be replaced by the value index particularly for problems optimizing these objectives. This reduces the process to achieve the best solutions and may find better classification for scenario definition. It is also concluded that decision makers are better to focus on value index and weighting its contents to find the most sustainable alternatives based on their requirements.

Keywords: Waste load allocation (WLA), Value index, Multi objective particle swarm optimization (MOPSO), Haraz River, Equity.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2027

Authors: Melis Inalpulat, Levent Genc

Abstract:

In the present study, an attempt was made to state the Land Use Land Cover (LULC) transformation over three decades around the urban regions of Balıkesir, Bursa, and Çanakkale provincial centers (PCs) in Turkey. Landsat imageries acquired in 1984, 1999 and 2014 were used to determine the LULC change. Images were classified using the supervised classification technique and five main LULC classes were considered including forest (F), agricultural land (A), residential area (urban) - bare soil (R-B), water surface (W), and other (O). Change detection analyses were conducted for 1984-1999 and 1999-2014, and the results were evaluated. Conversions of LULC types to R-B class were investigated. In addition, population changes (1985-2014) were assessed depending on census data, the relations between population and the urban areas were stated, and future populations and urban area needs were forecasted for 2030. The results of LULC analysis indicated that urban areas, which are covered under R-B class, were expanded in all PCs. During 1984-1999 R-B class within Balıkesir, Bursa and Çanakkale PCs were found to have increased by 7.1%, 8.4%, and 2.9%, respectively. The trend continued in the 1999-2014 term and the increment percentages reached to 15.7%, 15.5%, and 10.2% at the end of 30-year period (1984-2014). Furthermore, since A class in all provinces was found to be the principal contributor for the R-B class, urban sprawl lead to the loss of agricultural lands. Moreover, the areas of R-B classes were highly correlated with population within all PCs (R²>0.992). Depending on this situation, both future populations and R-B class areas were forecasted. The estimated values of increase in the R-B class areas for Balıkesir, Bursa, and Çanakkale PCs were 1,586 ha, 7,999 ha and 854 ha, respectively. Due to this fact, the forecasted values for 2,030 are 7,838 ha, 27,866, and 2,486 ha for Balıkesir, Bursa, and Çanakkale, and thus, 7.7%, 8.2%, and 9.7% more R-B class areas are expected to locate in PCs in respect to the same order.

Keywords: Landsat, LULC change, population, urban sprawl.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1437

Authors: Hadi Firouzi, Majid Nili Ahmadabadi, Babak N. Araabi

Abstract:

Conceptualization strengthens intelligent systems in generalization skill, effective knowledge representation, real-time inference, and managing uncertain and indefinite situations in addition to facilitating knowledge communication for learning agents situated in real world. Concept learning introduces a way of abstraction by which the continuous state is formed as entities called concepts which are connected to the action space and thus, they illustrate somehow the complex action space. Of computational concept learning approaches, action-based conceptualization is favored because of its simplicity and mirror neuron foundations in neuroscience. In this paper, a new biologically inspired concept learning approach based on the probabilistic framework is proposed. This approach exploits and extends the mirror neuron-s role in conceptualization for a reinforcement learning agent in nondeterministic environments. In the proposed method, instead of building a huge numerical knowledge, the concepts are learnt gradually from rewards through interaction with the environment. Moreover the probabilistic formation of the concepts is employed to deal with uncertain and dynamic nature of real problems in addition to the ability of generalization. These characteristics as a whole distinguish the proposed learning algorithm from both a pure classification algorithm and typical reinforcement learning. Simulation results show advantages of the proposed framework in terms of convergence speed as well as generalization and asymptotic behavior because of utilizing both success and failures attempts through received rewards. Experimental results, on the other hand, show the applicability and effectiveness of the proposed method in continuous and noisy environments for a real robotic task such as maze as well as the benefits of implementing an incremental learning scenario in artificial agents.

Keywords: Concept learning, probabilistic decision making, reinforcement learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1527

Authors: Abdulraaof H. Alqaili, Hamad A. Alsoliman

Abstract:

Through progress in pavement design developments, a pavement design method was developed, which is titled the Mechanistic Empirical Pavement Design Guide (MEPDG). Nowadays, the evolution in roads network and highways is observed in Saudi Arabia as a result of increasing in traffic volume. Therefore, the MEPDG currently is implemented for flexible pavement design by the Saudi Ministry of Transportation. Implementation of MEPDG for local pavement design requires the calibration of distress models under the local conditions (traffic, climate, and materials). This paper aims to prepare data for calibration of MEPDG in Central Saudi Arabia. Thus, the first goal is data collection for the design of flexible pavement from the local conditions of the Riyadh region. Since, the modifying of collected data to input data is needed; the main goal of this paper is the analysis of collected data. The data analysis in this paper includes processing each: Trucks Classification, Traffic Growth Factor, Annual Average Daily Truck Traffic (AADTT), Monthly Adjustment Factors (MAFi), Vehicle Class Distribution (VCD), Truck Hourly Distribution Factors, Axle Load Distribution Factors (ALDF), Number of axle types (single, tandem, and tridem) per truck class, cloud cover percent, and road sections selected for the local calibration. Detailed descriptions of input parameters are explained in this paper, which leads to providing of an approach for successful implementation of MEPDG. Local calibration of MEPDG to the conditions of Riyadh region can be performed based on the findings in this paper.

Keywords: Mechanistic-empirical pavement design guide, traffic characteristics, materials properties, climate, Riyadh.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1224

Authors: Siti Aisyah Zakaria, Nor Azrita Mohd Amin, Noor Fadhilah Ahmad Radi, Nasrul Hamidin

Abstract:

Higher ground-level ozone (GLO) concentration adversely affects human health, vegetations as well as activities in the ecosystem. In Malaysia, most of the analysis on GLO concentration are carried out using the average value of GLO concentration, which refers to the centre of distribution to make a prediction or estimation. However, analysis which focuses on the higher value or extreme value in GLO concentration is rarely explored. Hence, the objective of this study is to classify the tail behaviour of GLO using generalized extreme value (GEV) distribution estimation the return level using the corresponding modelling (Gumbel, Weibull, and Frechet) of GEV distribution. The results show that Weibull distribution which is also known as short tail distribution and considered as having less extreme behaviour is the best-fitted distribution for four selected air monitoring stations in Peninsular Malaysia, namely Larkin, Pelabuhan Kelang, Shah Alam, and Tanjung Malim; while Gumbel distribution which is considered as a medium tail distribution is the best-fitted distribution for Nilai station. The return level of GLO concentration in Shah Alam station is comparatively higher than other stations. Overall, return levels increase with increasing return periods but the increment depends on the type of the tail of GEV distribution’s tail. We conduct this study by using maximum likelihood estimation (MLE) method to estimate the parameters at four selected stations in Peninsular Malaysia. Next, the validation for the fitted block maxima series to GEV distribution is performed using probability plot, quantile plot and likelihood ratio test. Profile likelihood confidence interval is tested to verify the type of GEV distribution. These results are important as a guide for early notification on future extreme ozone events.

Keywords: Extreme value theory, generalized extreme value distribution, ground-level ozone, return level.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 519

Authors: Hamidullah Binol, Abdullah Bal

Abstract:

Nowadays, food safety is a great public concern; therefore, robust and effective techniques are required for detecting the safety situation of goods. Hyperspectral Imaging (HSI) is an attractive material for researchers to inspect food quality and safety estimation such as meat quality assessment, automated poultry carcass inspection, quality evaluation of fish, bruise detection of apples, quality analysis and grading of citrus fruits, bruise detection of strawberry, visualization of sugar distribution of melons, measuring ripening of tomatoes, defect detection of pickling cucumber, and classification of wheat kernels. HSI can be used to concurrently collect large amounts of spatial and spectral data on the objects being observed. This technique yields with exceptional detection skills, which otherwise cannot be achieved with either imaging or spectroscopy alone. This paper presents a nonlinear technique based on kernel Fukunaga-Koontz transform (KFKT) for detection of fat content in ground meat using HSI. The KFKT which is the nonlinear version of FKT is one of the most effective techniques for solving problems involving two-pattern nature. The conventional FKT method has been improved with kernel machines for increasing the nonlinear discrimination ability and capturing higher order of statistics of data. The proposed approach in this paper aims to segment the fat content of the ground meat by regarding the fat as target class which is tried to be separated from the remaining classes (as clutter). We have applied the KFKT on visible and nearinfrared (VNIR) hyperspectral images of ground meat to determine fat percentage. The experimental studies indicate that the proposed technique produces high detection performance for fat ratio in ground meat.

Keywords: Food (Ground meat) inspection, Fukunaga-Koontz transform, hyperspectral imaging, kernel methods.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1500

Authors: Mark Joseph Quinto, Roan Beronilla, Guiller Damian, Eliza Camaso, Ronaldo Alberto

Abstract:

Forest inventories are essential to assess the composition, structure and distribution of forest vegetation that can be used as baseline information for management decisions. Classical forest inventory is labor intensive and time-consuming and sometimes even dangerous. The use of Light Detection and Ranging (LiDAR) in forest inventory would improve and overcome these restrictions. This study was conducted to determine the possibility of using LiDAR derived data in extracting high accuracy forest biophysical parameters and as a non-destructive method for forest status analysis of San Manual, Pangasinan. Forest resources extraction was carried out using LAS tools, GIS, Envi and .bat scripts with the available LiDAR data. The process includes the generation of derivatives such as Digital Terrain Model (DTM), Canopy Height Model (CHM) and Canopy Cover Model (CCM) in .bat scripts followed by the generation of 17 composite bands to be used in the extraction of forest classification covers using ENVI 4.8 and GIS software. The Diameter in Breast Height (DBH), Above Ground Biomass (AGB) and Carbon Stock (CS) were estimated for each classified forest cover and Tree Count Extraction was carried out using GIS. Subsequently, field validation was conducted for accuracy assessment. Results showed that the forest of San Manuel has 73% Forest Cover, which is relatively much higher as compared to the 10% canopy cover requirement. On the extracted canopy height, 80% of the tree’s height ranges from 12 m to 17 m. CS of the three forest covers based on the AGB were: 20819.59 kg/20x20 m for closed broadleaf, 8609.82 kg/20x20 m for broadleaf plantation and 15545.57 kg/20x20m for open broadleaf. Average tree counts for the tree forest plantation was 413 trees/ha. As such, the forest of San Manuel has high percent forest cover and high CS.

Keywords: Carbon stock, forest inventory, LiDAR, tree count.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1281

Authors: B. L. Gadiga

Abstract:

This study assesses the major ecological zones in Nigeria with the view to understanding the spatial pattern of vegetation zones and the implications on conservation within the period of sixteen (16) years. Satellite images used for this study were acquired from the SPOT-VEGETATION between 1998 and 2013. The annual NDVI images selected for this study were derived from SPOT-4 sensor and were acquired within the same season (November) in order to reduce differences in spectral reflectance due to seasonal variations. The images were sliced into five classes based on literatures and knowledge of the area (i.e. <0.16 Non-Vegetated areas; 0.16-0.22 Sahel Savannah; 0.22-0.40 Sudan Savannah, 0.40-0.47 Guinea Savannah and >0.47 Forest Zone). Classification of the 1998 and 2013 images into forested and non forested areas showed that forested area decrease from 511,691 km² in 1998 to 478,360 km² in 2013. Differencing change detection method was performed on 1998 and 2013 NDVI images to identify areas of ecological concern. The result shows that areas undergoing vegetation degradation covers an area of 73,062 km² while areas witnessing some form restoration cover an area of 86,315 km². The result also shows that there is a weak correlation between rainfall and the vegetation zones. The non-vegetated areas have a correlation coefficient (r) of 0.0088, Sahel Savannah belt 0.1988, Sudan Savannah belt -0.3343, Guinea Savannah belt 0.0328 and Forest belt 0.2635. The low correlation can be associated with the encroachment of the Sudan Savannah belt into the forest belt of South-eastern part of the country as revealed by the image analysis. The degradation of the forest vegetation is therefore responsible for the serious erosion problems witnessed in the South-east. The study recommends constant monitoring of vegetation and strict enforcement of environmental laws in the country.

Keywords: Vegetation, NDVI, SPOT-vegetation, ecology, degradation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 839

Authors: Peng Zhang, Lihui Zheng, Xiangchun Wang, Xiaopan Kou

Abstract:

Quantitative research on the main control factors of lost circulation has few considerations and single data source. Using Unmanned Intervention Algorithm to find the main control factors of lost circulation adopts all measurable parameters. The degree of lost circulation is characterized by the loss rate as the objective function. Geological, engineering and fluid data are used as layers, and 27 factors such as wellhead coordinates and Weight on Bit (WOB) used as dimensions. Data classification is implemented to determine function independent variables. The mathematical equation of loss rate and 27 influencing factors is established by multiple regression method, and the undetermined coefficient method is used to solve the undetermined coefficient of the equation. Only three factors in t-test are greater than the test value 40, and the F-test value is 96.557%, indicating that the correlation of the model is good. The funnel viscosity, final shear force and drilling time were selected as the main control factors by elimination method, contribution rate method and functional method. The calculated values of the two wells used for verification differ from the actual values by -3.036 m3/h and -2.374 m3/h, with errors of 7.21% and 6.35%. The influence of engineering factors on the loss rate is greater than that of funnel viscosity and final shear force, and the influence of the three factors is less than that of geological factors. The best combination of funnel viscosity, final shear force and drilling time is obtained through quantitative calculation. The minimum loss rate of lost circulation wells in Shunbei area is 10 m3/h. It can be seen that man-made main control factors can only slow down the leakage, but cannot fundamentally eliminate it. This is more in line with the characteristics of karst caves and fractures in Shunbei fault solution oil and gas reservoir.

Keywords: Drilling fluid, loss rate, main controlling factors, Unmanned Intervention Algorithm.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 403

Authors: Mirza Munir Ahmed, S.R.M. Kutty, Mohd Faris Khamidi, Idris Othman, Azmi Mohd Shariff

Abstract:

Petrol Fuel Station (PFS) has potential hazards to the people, asset, environment and reputation of an operating company. Fire hazards, static electricity air pollution evoked by aliphatic and aromatic organic compounds are major causes of accident/incident occurrence at fuel station. Activities such as carelessness, maintenance, housekeeping, slips trips and falls, transportation hazard, major and minor injuries, robbery and snake bites has a potential to create unsafe conditions. The level of risk of these hazards varies according to location and country. The emphasis on safety considerations by the government is variable all around the world. Developed countries safety records are much better as compared to developing countries safety statistics. There is no significant approach available to highlight the unsafe acts and unsafe conditions during operation and maintenance of fuel station. Fuel station is the most commonly available facilities that contain flammable and hazardous materials. Due to continuous operation of fuel station they pose various hazards to people, environment and assets of an organization. To control these hazards, there is a need for specific approach. PFS operation is unique as compared to other businesses. For smooth operations it demands an involvement of operating company, contractor and operator group. This study will focus to address hazard contributing factors that have a potential to make PFS operation risky. One year data collected, 902 activities analyzed, comparisons were made to highlight significant contributing factors. The study will provide help and assistance to PFS outlet marketing companies to make their fuel station operation safer. It will help health safety and environment (HSE) professionals to arrest the gap available related to safety matters at PFS.

Keywords: Accident, Contributing factors, carelessness, fire, explosion, injuries.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 7385

Authors: Ticijana Prijon, Branko Ermenc

Abstract:

Background: Blunt aortic trauma (BAT) includes various morphological changes that occur during deceleration, acceleration and/or body compression in traffic accidents. The various forms of BAT, from limited laceration of the intima to complete transection of the aorta, depends on the force acting on the vessel wall and the tolerance of the aorta to injury. The force depends on the change in velocity, the dynamics of the accident and of the seating position in the car. Tolerance to aortic injury depends on the anatomy, histological structure and pathomorphological alterations due to aging or disease of the aortic wall. An overview of the literature and medical documentation reveals that different terms are used to describe certain forms of BAT, which can lead to misinterpretation of findings or diagnoses. We therefore, propose a classification that would enable uniform systematic screening of all forms of BAT. We have classified BAT into three morphologycal types: TYPE I (intramural), TYPE II (transmural) and TYPE III (multiple) aortic ruptures with appropriate subtypes. Methods: All car accident casualties examined at the Institute of Forensic Medicine from 2001 to 2009 were included in this retrospective study. Autopsy reports were used to determine the occurrence of each morphological type of BAT in deceased drivers, front seat passengers and other passengers in cars and to define the morphology of BAT in relation to the accident dynamics and the age of the fatalities. Results: A total of 391 fatalities in car accidents were included in the study. TYPE I, TYPE II and TYPE III BAT were observed in 10,9%, 55,6% and 33,5%, respectively. The incidence of BAT in drivers, front seat and other passengers was 36,7%, 43,1% and 28,6%, respectively. In frontal collisions, the incidence of BAT was 32,7%, in lateral collisions 54,2%, and in other traffic accidents 29,3%. The average age of fatalities with BAT was 42,8 years and of those without BAT 39,1 years. Conclusion: Identification and early recognition of the risk factors of BAT following a traffic accident is crucial for successful treatment of patients with BAT. Front seat passengers over 50 years of age who have been injured in a lateral collision are the most at risk of BAT.

Keywords: Aorta, blunt trauma, car accidents, morphology, risk factors.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2104

Authors: Kaveh Mollazade, Hojat Ahmadi, Mahmoud Omid, Reza Alimardani

Abstract:

Recently, the issue of machine condition monitoring and fault diagnosis as a part of maintenance system became global due to the potential advantages to be gained from reduced maintenance costs, improved productivity and increased machine availability. The aim of this work is to investigate the effectiveness of a new fault diagnosis method based on power spectral density (PSD) of vibration signals in combination with decision trees and fuzzy inference system (FIS). To this end, a series of studies was conducted on an external gear hydraulic pump. After a test under normal condition, a number of different machine defect conditions were introduced for three working levels of pump speed (1000, 1500, and 2000 rpm), corresponding to (i) Journal-bearing with inner face wear (BIFW), (ii) Gear with tooth face wear (GTFW), and (iii) Journal-bearing with inner face wear plus Gear with tooth face wear (B&GW). The features of PSD values of vibration signal were extracted using descriptive statistical parameters. J48 algorithm is used as a feature selection procedure to select pertinent features from data set. The output of J48 algorithm was employed to produce the crisp if-then rule and membership function sets. The structure of FIS classifier was then defined based on the crisp sets. In order to evaluate the proposed PSD-J48-FIS model, the data sets obtained from vibration signals of the pump were used. Results showed that the total classification accuracy for 1000, 1500, and 2000 rpm conditions were 96.42%, 100%, and 96.42% respectively. The results indicate that the combined PSD-J48-FIS model has the potential for fault diagnosis of hydraulic pumps.

Keywords: Power Spectral Density, Machine ConditionMonitoring, Hydraulic Pump, Fuzzy Logic.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2713

Authors: Sokkhey Phauk, Takeo Okazaki

Abstract:

The challenging task in educational institutions is to maximize the high performance of students and minimize the failure rate of poor-performing students. An effective method to leverage this task is to know student learning patterns with highly influencing factors and get an early prediction of student learning outcomes at the timely stage for setting up policies for improvement. Educational data mining (EDM) is an emerging disciplinary field of data mining, statistics, and machine learning concerned with extracting useful knowledge and information for the sake of improvement and development in the education environment. The study is of this work is to propose techniques in EDM and integrate it into a web-based system for predicting poor-performing students. A comparative study of prediction models is conducted. Subsequently, high performing models are developed to get higher performance. The hybrid random forest (Hybrid RF) produces the most successful classification. For the context of intervention and improving the learning outcomes, a feature selection method MICHI, which is the combination of mutual information (MI) and chi-square (CHI) algorithms based on the ranked feature scores, is introduced to select a dominant feature set that improves the performance of prediction and uses the obtained dominant set as information for intervention. By using the proposed techniques of EDM, an academic performance prediction system (APPS) is subsequently developed for educational stockholders to get an early prediction of student learning outcomes for timely intervention. Experimental outcomes and evaluation surveys report the effectiveness and usefulness of the developed system. The system is used to help educational stakeholders and related individuals for intervening and improving student performance.

Keywords: Academic performance prediction system, prediction model, educational data mining, dominant factors, feature selection methods, student performance.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 980

Authors: Rashima Mahajan, Dipali Bansal, Shweta Singh

Abstract:

Real time non-invasive Brain Computer Interfaces have a significant progressive role in restoring or maintaining a quality life for medically challenged people. This manuscript provides a comprehensive review of emerging research in the field of cognitive/affective computing in context of human neural responses. The perspectives of different emotion assessment modalities like face expressions, speech, text, gestures, and human physiological responses have also been discussed. Focus has been paid to explore the ability of EEG (Electroencephalogram) signals to portray thoughts, feelings, and unspoken words. An automated workflow-based protocol to design an EEG-based real time Brain Computer Interface system for analysis and classification of human emotions elicited by external audio/visual stimuli has been proposed. The front end hardware includes a cost effective and portable Emotiv EEG Neuroheadset unit, a personal computer and a set of external stimulators. Primary signal analysis and processing of real time acquired EEG shall be performed using MATLAB based advanced brain mapping toolbox EEGLab/BCILab. This shall be followed by the development of MATLAB based self-defined algorithm to capture and characterize temporal and spectral variations in EEG under emotional stimulations. The extracted hybrid feature set shall be used to classify emotional states using artificial intelligence tools like Artificial Neural Network. The final system would result in an inexpensive, portable and more intuitive Brain Computer Interface in real time scenario to control prosthetic devices by translating different brain states into operative control signals.

Keywords: Brain Computer Interface (BCI), Electroencephalogram (EEG), EEGLab, BCILab, Emotiv, Emotions, Interval features, Spectral features, Artificial Neural Network, Control applications.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 5297

Authors: Mustafa Al Sukar, Azzam Sleit, Abdullatif Abu-Dalhoum, Bassam Al-Kasasbeh

Abstract:

Use and abuse of drugs by teens is very common and can have dangerous consequences. The drugs contribute to physical and sexual aggression such as assault or rape. Some teenagers regularly use drugs to compensate for depression, anxiety or a lack of positive social skills. Teen resort to smoking should not be minimized because it can be "gateway drugs" for other drugs (marijuana, cocaine, hallucinogens, inhalants, and heroin). The combination of teenagers' curiosity, risk taking behavior, and social pressure make it very difficult to say no. This leads most teenagers to the questions: "Will it hurt to try once?" Nowadays, technological advances are changing our lives very rapidly and adding a lot of technologies that help us to track the risk of drug abuse such as smart phones, Wireless Sensor Networks (WSNs), Internet of Things (IoT), etc. This technique may help us to early discovery of drug abuse in order to prevent an aggravation of the influence of drugs on the abuser. In this paper, we have developed a Decision Support System (DSS) for detecting the drug abuse using Artificial Neural Network (ANN); we used a Multilayer Perceptron (MLP) feed-forward neural network in developing the system. The input layer includes 50 variables while the output layer contains one neuron which indicates whether the person is a drug addict. An iterative process is used to determine the number of hidden layers and the number of neurons in each one. We used multiple experiment models that have been completed with Log-Sigmoid transfer function. Particularly, 10-fold cross validation schemes are used to access the generalization of the proposed system. The experiment results have obtained 98.42% classification accuracy for correct diagnosis in our system. The data had been taken from 184 cases in Jordan according to a set of questions compiled from Specialists, and data have been obtained through the families of drug abusers.

Keywords: Artificial Neural Network, Decision Support System, drug abuse, drug addiction, Multilayer Perceptron.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1682

Authors: Ebipatei Victoria Tunyan, T. A. Cao, Cheol Young Ock

Abstract:

Detecting subjectively biased statements is a vital task. This is because this kind of bias, when present in the text or other forms of information dissemination media such as news, social media, scientific texts, and encyclopedias, can weaken trust in the information and stir conflicts amongst consumers. Subjective bias detection is also critical for many Natural Language Processing (NLP) tasks like sentiment analysis, opinion identification, and bias neutralization. Having a system that can adequately detect subjectivity in text will boost research in the above-mentioned areas significantly. It can also come in handy for platforms like Wikipedia, where the use of neutral language is of importance. The goal of this work is to identify the subjectively biased language in text on a sentence level. With machine learning, we can solve complex AI problems, making it a good fit for the problem of subjective bias detection. A key step in this approach is to train a classifier based on BERT (Bidirectional Encoder Representations from Transformers) as upstream model. BERT by itself can be used as a classifier; however, in this study, we use BERT as data preprocessor as well as an embedding generator for a Bi-LSTM (Bidirectional Long Short-Term Memory) network incorporated with attention mechanism. This approach produces a deeper and better classifier. We evaluate the effectiveness of our model using the Wiki Neutrality Corpus (WNC), which was compiled from Wikipedia edits that removed various biased instances from sentences as a benchmark dataset, with which we also compare our model to existing approaches. Experimental analysis indicates an improved performance, as our model achieved state-of-the-art accuracy in detecting subjective bias. This study focuses on the English language, but the model can be fine-tuned to accommodate other languages.

Keywords: Subjective bias detection, machine learning, BERT–BiLSTM–Attention, text classification, natural language processing.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 830