LDA Method Research Articles

Sentiment Analysis And Topic Modeling on User Reviews of Online Tutoring Applications Using Support Vector Machine and Latent Dirichlet Allocation

<p>Ruangguru is an online non-formal education application in Indonesia. There are several appealing features that encourage students to study online. The app's release on the Google Play Store will assist app developers in receiving feedback through the review feature.Users submit various topics and comments about Ruangguru in the review feature of Ruangguru, making it difficult to manually identify public sentiments and topics of conversation. Opinions submitted by users on the review feature are interesting to research further. This study aims to classify user opinions into positive and negative classes and model topics in both classes. Topic modeling aims to find out the topics that are often discussed in each class. The stages of this study include data collection, data cleaning, data transformation, and data classification with the Support Vector Machine method and the Latent Dirichlet Allocation method for topic modeling. The results of topic modeling with the LDA method in each positive and negative class can be seen from the coherence value. Namely, the higher the coherence value of a topic, the easier the topic is interpreted by humans. The testing process in this study used Confusion Matrix and ROUGE. The results of model performance testing using the Confusion Matrix are shown with accuracy, precision, recall, and f-measure values of 0.9, 0.9, 0.9, and 0.89, respectively. The results of model performance testing using ROUGE resulted in the highest recall, precision, and f-measure of 1, 0.84, and 0.91. The highest coherence value is found in the 20th topic, with a value of 0.318. Using the Support Vector Machine and Latent Dirichlet Allocation algorithms are considered adequate for sentiment analysis and topic modeling for the Ruangguru application.</p>

Comparative study of LDA and PIV methods for flow homogeneity measurements

Abstract In the Hydraulic Laboratory of the Department of Hydraulic Engineering, Faculty of Civil Engineering, Slovak University of Technology in Bratislava the distribution of the flow velocity in a shaft intake structure was investigated on a physical model. The homogeneity of the flow velocity distribution in the intakes of turbines has a great influence on the turbine performance. Modern optical methods, particle image velocimetry and laser Doppler anemometry, were used to measure the flow velocity distribution in the model of the turbine intake. The particle image velocimetry method provides a record of the velocity field, while the laser Doppler anemometry method provides an accurate indication of the velocity at a given point. The flow velocity distributions in the examined profiles were measured by the particle image velocimetry and laser Doppler anemometry methods and the results were compared with each other.

Spatial and sentiment analysis of public opinion toward COVID-19 pandemic using twitter data: At the early stage of vaccination

During the crisis of Coronavirus pandemic, social media, like Twitter, have been the platforms on which people have been able to share their opinions and obtain information. The present study provides a detailed spatial-temporal analysis of the Twitter online discourse (approximately 280 thousand tweets) in Ohio and Michigan at the early stage of vaccination rollout (January 2021, till March 2021). This work aims to explore how people were feeling about the pandemic, the most frequent topics people were talking about, and how the topics spatially were distributed. Moreover, state government responses and important news were gathered to analyze their impacts on public opinion based on the temporal analysis of the tweets. In this project, Natural Language Processing using the LDA method was employed to identify 11 topics and 8 sub-topics in the Twitter data. The temporal analysis of topics shows the sensitivity of the online discourse to the significant state news and the local government's reactions to the pandemic. Moreover, the spatial distribution of Coronavirus-related tweets and sentiments demonstrates concentrations in the more populated urban areas with a high rate of COVID-19 cases in Ohio and Michigan. The government's economic and financial policies taken during this time, the vaccination timeline phases specified by each state, and the pandemic-related information can contribute to public opinion and sentiment trends. The findings of this study can help explore public demands, and reactions, follow the impacts of the local authorities' policies at the county level and manage their future responses to such a pandemic.

Preliminary study non-destructive sorting techniques for pepper (Capsicum annuum L.) using odor parameter

Almost all fruits and vegetables sold in modern society are sorted and labeled, making it easier for customers to recognize the quality of the product, leading to more regular distribution and supply. Consequently, it facilitates the initial packaging and transportation of the product, and farmers will benefit from the added value. Therefore, it is necessary to develop sorting via affordable machines and easy to operate at the current technology level. Since electronic nose technology is new-emerging, it can be used in food quality control systems. In this study, the variety Padrón (Capsicum annuum L.) was evaluated. PCA, LDA, SVM and ANN methods were used to classify sweet and hot peppers. According to PCA, 98% of the variance in the data was detected by the first three components. SVM, ANN, and LDA all showed 100% accuracy in classification. The amounts of capsaicin in two types of sweet and hot peppers were predicted well and with high accuracy by three different methods: MLR, PCR, and PLSR. With this method, it is possible to reliably separate sweet and hot peppers based on odor parameters, and it is also possible to develop sorting machines according to the characteristics of odor.

토픽모델링을 통한 우수 원격수업 특성 및 요소 연구

Objectives The purposes of this study is to identify the characteristics and elements of exemplary remote learning, provide factors that control the quality of effective remote learning, and present implications according to the educational environment changed by the COVID-19 situation. Methods For this study, a total of 324 students(1,178 case) and 459 instructors(459 case) in University A located in Seoul were surveyed using Google Survey, and the collected data was analyzed through topic modelling’s LDA method. Results Survey results show that instructors and students each perceived four topics of remote learning elements. Topics extracted from the students’ surveys are “prompt response to problems that occur”, “difference in the quality of lectures according to the instructors’ design of remote lectures”, “systematic preparation of remote learning and good communication”, and “recreating a curriculum that is suitable for remote learning”. Topics extracted from the instructors’ survey are “supporting the environment and equipment needed for remote learning”, “specific guidelines for remote learning”, “provision of clear grading guideline for remote learning”, and “good communication in remote learning.” Conclusions The results of this study present practical implications for smooth remote learning, such as the need for both students and instructors to have basic media literacy necessary for remote learning and the need to redesign teaching and learning to fit the remote learning environment(e.g., systematic redesign of lectures, interaction, class evaluation criteria, etc., appropriate for remote learning), and propose future research topics.

Quality Service Improvement using Hybrid Big Data Analytics Model: A Case of AYCE Korean BBQ Restaurant

The culinary industry continues to develop every year. One of the phenomena is all you can eat (AYCE) Korean BBQ restaurant. This will cause increasingly intense competition in the restaurant business. The aims of this study are to identify customer needs, to determine the priority of the service attributes, and to provide appropriate suggestions to AYCE Korean BBQ restaurant. The big data analysis using LDA method was applied to determine customer needs based on the Zomato review. The results showed that ten customer needs are identified as Voice of Customer (VOC) for HOQ of the Quality Function Deployment (QFD). The results of the AHP showed that 5 priorities of improvement strategies. The appropriate suggestions for Korean BBQ restaurant are to create a layered control system and customer satisfaction surveys, to implement FIFO and LIFO systems, to check the condition of raw material storage equipment, to improve existing service SOPs, and to provide periodic briefings and training.

Экологическая проблематика в предвыборных программах парламентских партий России в 2021 г.

The article presents the relationship between the environmental themes of social media and the election programs of the parliamentary parties. The author analyzes the relevance of the environmental agenda in modern Russia and identifies two multidirectional vectors in the environmental sphere. One vector is the problems and cases that require prompt solutions, the other – problems and cases requiring strategic solutions. Based on the analysis of user comments under posts on the topic of ecology in social media VKontakte, Odnoklassniki, Instagram, Facebook, Twitter, the paper identify six main environmental topics using thematic modeling by the LDA method and 35 points on environmental topics based on the analysis of the election programs of five parliamentary parties. The author identifies 19 points of election programs directly or indirectly intersecting with six environmental topics of social networks. The most complete environmental topics are presented in the United Russia program, which shows the largest number of intersections with environmental topics of social media.

Spatio-Temporal Patterns of Fitness Behavior in Beijing Based on Social Media Data

Fitness is an important way to ensure the health of the population, and it is important to actively understand fitness behavior. Although social media Weibo data (the Chinese Tweeter) can provide multidimensional information in terms of objectivity and generalizability, there is still more latent potential to tap. Based on Sina Weibo social media data in the year 2017, this study was conducted to explore the spatial and temporal patterns of urban residents’ different fitness behaviors and related influencing factors within the Fifth Ring Road of Beijing. FastAI, LDA, geodetector technology, and GIS spatial analysis methods were employed in this study. It was found that fitness behaviors in the study area could be categorized into four types. Residents can obtain better fitness experiences in sports venues. Different fitness types have different polycentric spatial distribution patterns. The residents’ fitness frequency shows an obvious periodic distribution (weekly and 24 h). The spatial distribution of the fitness behavior of residents is mainly affected by factors, such as catering services, education and culture, companies, and public facilities. This research could help to promote the development of urban residents’ fitness in Beijing.

Classification of hyperspectral remote sensing images using different dimension reduction methods with 3D/2D CNN

The high dimensionality of hyperspectral remote sensing images (HRSI) affects the classification performance. Therefore, most HRSI classification methods use dimension reduction methods as a solution for high dimensionality. It is aimed to extract useful features with dimension reduction methods. At the end of this process, the data dimension is reduced and the transaction cost is decreased. In this study, LDA, PCA, IPCA, ICA, SPCA, RPCA and SVD dimension reduction methods were applied as a preprocessing step to improve HRSI classification performance. Since HRSI is volumetric data and has a spectral dimension, 2D CNN cannot extract good distinguishing features from spectral dimensions. Because 2D CNN only considers spatial information. With 3D CNN, spectral-spatial features are extracted simultaneously. However, 3D CNN increases the computational cost. Therefore, in this study, Hybrid 3D/2D CNN method is used together with dimension reduction methods. Hybrid CNN method consists of a combination of 3D CNN, 2D CNN and depthwise separable convolution. While 3D CNN extracts common spectral-spatial features, more spatial features are learned with 2D CNN used after 3D CNN. With depthwise separable convolution, it reduces the number of parameters and overfitting is prevented. The applications performed on the frequently used HRSI benchmark datasets show that the classification performance of the proposed method is better than the compared methods. In addition, Indian pines, HyRANK-Loukia, Botswana and Pavia of University datasets are used to examine the effect of dimension reduction methods used together with the hybrid 3D/2D CNN method on classification performance. As a result of the applications, the best classification accuracies were obtained in PCA, LDA and IPCA with Indian pines, PCA with Pavia of university, PCA and IPCA with Salinas, PCA, RPCA and LDA dimension reduction methods with HyRANK-Loukia.

Construction of a Legal System of Corporate Social Responsibility Based on Big Data Analysis Technology.

The company is an essential organization in modern society, and the company has transformed from a purely economic organization to a corporate citizen that realizes economic responsibility and practices social responsibility at the same time. It is only by constructing a legal system of corporate social responsibility that companies can take social responsibility on the track of the legal system, realize the company's mission of the times, and achieve a win-win situation for both the company and society. This paper used the LDA and text clustering methods to analyze existing legal texts. It obtained the theme and text clustering results, thus proposing five aspects of the legal system construction framework to guide the corporate social responsibility legal system, which has pioneering significance.

Feasibility of Visible/Near Infrared (Vis/NIR) Spectrophotometry capability in classification of lemon samples during storage period by PCA, LDA and SVM identification methods

Feasibility of Visible/Near Infrared (Vis/NIR) Spectrophotometry capability in classification of lemon samples during storage period by PCA, LDA and SVM identification methods

View analysis of personal information leakage and privacy protection in big data era—based on Q method

PurposeIn the era of big data, people are more likely to pay attention to privacy protection with facing the risk of personal information leakage while enjoying the convenience brought by big data technology. Furthermore, people’s views on personal information leakage and privacy protection are varied, playing an important role in the legal process of personal information protection. Therefore, this paper aims to propose a semi-qualitative method based framework to reveal the subjective patterns about information leakage and privacy protection and further provide practical implications for interested party.Design/methodology/approachQ method is a semi-qualitative methodology which is designed for identifying typologies of perspectives. In order to have a comprehensive understanding of users’ viewpoints, this study incorporates LDA & TextRank method and other information extraction technologies to capture the statements from large-scale literature, app reviews, typical cases and survey interviews, which could be regarded as the resource of the viewpoints.FindingsBy adopting the Q method that aims for studying subjective thought patterns to identify users’ potential views, the authors have identified three categories of stakeholders’ subjectivities: macro-policy sensitive, trade-offs and personal information sensitive, each of which perceives different risk and affordance of information leakage and importance and urgency of privacy protection. All of the subjectivities of the respondents reflect the awareness of the issue of information leakage, that is, the interested parties like social network sites are unable to protect their full personal information, while reflecting varied resistance and susceptibility of disclosing personal information for big data technology applications.Originality/valueThe findings of this study provide an overview of the subjective patterns on the information leakage issue. Being the first to incorporate the Q method to study the views of personal information leakage and privacy protection, the research not only broadens the application field of the Q method but also enriches the research methods for personal information protection. Besides, the proposed LDA & TextRank method in this paper alleviates the limitation of statements resource in the Q method.

Rapid discrimination of Curcuma longa and Curcuma xanthorrhiza using Direct Analysis in Real Time Mass Spectrometry and Near Infrared Spectroscopy

This study describes a newly developed method for the fast and straightforward differentiation of two turmeric species using Direct Analysis in Real Time mass spectrometry and miniaturized Near Infrared spectroscopy. Multivariate analyses (PCA and LDA) were performed on the mass spectrometric data, thus creating a powerful model for the discrimination of Curcumalonga and Curcumaxanthorrhiza. Cross-validation of the model revealed correctness-scores of 100% with 20-fold as well as leave-one-out validation techniques. To further estimate the models prediction power, seven retail samples of turmeric powder were analyzed and assorted to a species. Looking for a fast, non-invasive, cost-efficient and laboratory independent method, miniaturized NIR spectrometers offer an alternative for quality control of turmeric species. However, different technologies implemented to compensate for their small size, lead to different applicability of these spectrometers. Therefore, we investigated the three handheld spectrometers microPHAZIR, MicroNIR 2200 and MicroNIR 1700ES for their application in spice analysis in hyphenation to PCA, LDA and ANN methods used for the discriminant analysis. While microPHAZIR proved to be the most valuable device for differentiating C.longa and C.xanthorrhiza, MicroNIR 1700ES offered the worst results. These findings are interpreted on the basis of a quantum chemical simulation of the NIR spectrum of curcumin as the representative constituent. It was found that the information accessible to MicroNIR 1700ES that is relevant to the analyzed constituents is located in the spectral region prone to interferences with the matrix, likely limiting the performance of this spectrometer in this analytical scenario.

A Machine Learning Method for Classification and Identification of Potato Cultivars Based on the Reaction of MOS Type Sensor-Array.

In response to one of the most important challenges of the century, i.e., the estimation of the food demands of a growing population, advanced technologies have been employed in agriculture. The potato has the main contribution to people’s diet worldwide. Therefore, its different aspects are worth studying. The large number of potato varieties, lack of awareness about its new cultivars among farmers to cultivate, time-consuming and inaccurate process of identifying different potato cultivars, and the significance of identifying potato cultivars and other agricultural products (in every food industry process) all necessitate new, fast, and accurate methods. The aim of this study was to use an electronic nose, along with chemometrics methods, including PCA, LDA, and ANN as fast, inexpensive, and non-destructive methods for detecting different potato cultivars. In the present study, nine sensors with the best response to VOCs were adopted. VOCs sensors were used at various VOCs concentrations (1 to 10,000 ppm) to detect different gases. The results showed that a PCA with two main components, PC1 and PC2, described 92% of the total samples’ dataset variance. In addition, the accuracy of the LDA and ANN methods were 100 and 96%, respectively.

Sentiment Analysis of Public Opinion Related to Rapid Test Using LDA Method

In 2020 the world will be shocked by an outbreak of a disease that has developed tremendously. This disease is the Coronavirus. The Indonesian government, in overcoming conducted a Rapid early detection test in the spread of the Coronavirus. The steps of the Indonesian government have received rejection in several areas because people consume hoax news on social media. Indonesians widely use Twitter in conversations about the Coronavirus. Previous research was carried out using large-scale data, which affected the performance of the topic extraction method. The classification used resulted in poor accuracy using LDA to find the probability of topics in existing documents. LDA excels in large-scale data processing and is more consistent in generating the topic proportion value and word probability. Aspect-based sentiment analysis on public opinion regarding the rapid test on Twitter using LDA can determine aspects and public opinion on the rapid test. The test results of this study obtained 7000 tweets, four aspects of the results of topic using LDA, and getting the best accuracy using the RBF kernel by 95%. The sentiment of the Indonesian people towards the Rapid test is positive, with 4,305 sentiments.

Detection of the bacteria concentration level in pasteurized milk by using two different artificial multisensory methods

The purpose of this paper is to describe the use of an E-nose and E-tongue that were evaluated for the Escherichia coli detection at different concentrations and their ability to discriminate this bacterium from others, such as Klebsiella pneumoniae and Salmonella enterica in pasteurized milk.For data processing, the PCA and LDA methods were applied. Likewise, for the data classification, the SVM and k-NN methods were used. Moreover, each method was applied to the data set obtained by both sensory systems which had different data dimensionality.For detecting and classifying E. coli, S. enterica, and K. pneumoniae in pasteurized milk, it was observed that both systems obtained comparable results with 94.7% and a 92.5% success rate. Thus the devices successfully detected and classified the three bacteria tested, clearly differentiating them from the sterile milk samples. On the other hand, the E-tongue with the gold electrode achieved a 98.7% success rate in the discrimination of decreasing concentrations of E. coli, from 1 × 106 CFU/mL to 1 × 10−2 CFU/mL, in pasteurized milk.

How do developers discuss and support new programming languages in technical Q&A site? An empirical study of Go, Swift, and Rust in Stack Overflow

New programming languages (e.g., Swift, Go, Rust, etc.) are being introduced to provide a better opportunity for the developers to make software development robust and easy. At the early stage, a programming language is likely to have resource constraints that encourage the developers to seek help frequently from experienced peers active in QA sites such as Stack Overflow (SO). In this study, we have formally studied the discussions on three popular new languages introduced after the inception of SO (2008) and match those with the relevant activities in GitHub whenever appropriate. For that purpose, we have mined 4,17,82,536 questions and answers from SO and 7,846 issue information along with 6,60,965 repository information from GitHub. Initially, the development of new languages is relatively slow compared to mature languages (e.g., C, C++, Java). The expected outcome of this study is to reveal the difficulties and challenges faced by the developers working with these languages so that appropriate measures can be taken to expedite the generation of relevant resources. We have used the LDA method on SO's questions and answers to identify different topics of new languages. We have extracted several features of the answer pattern of the new languages from SO to study their characteristics. These attributes were used to identify difficult topics. We explored the background of developers who are contributing to these languages. We have created a model by combining Stack Overflow data and issues, repository, user data of GitHub. Finally, we have used that model to identify factors that affect language evolution. We believe that the outcome of our study is likely to help the owner/sponsor of these languages to design better features and documentation. It will also help the software developers or students to prepare themselves to work on these languages in an informed way.

Application of electronic nose with chemometrics methods to the detection of juices fraud

Today, consumer demands for fresh foods with desirable sensory properties are increasing. Lack of easy, cheap, and nondestructive methods to control the quality of fruit juices is one of the main challenges in the beverage industry. The present study has examined two types of natural and industrial juices. Machine olfaction with 9-metal oxide semiconductor (MOS) sensors was used for the experiments. Sensor response patterns were analyzed using chemometrics methods, including principal component analysis (PCA), linear and quadratic discriminant analyses (LDA and QDA), and support vector machine (SVM). According to the results, the classification accuracy of SVM, QDA, and LDA methods was 96.25%, 95.8%, and 94.4%, respectively. Practical applications It is important to note that most fruit juices sold in supermarkets contain only a small percentage of natural fruit juice and usually include added sweeteners. As a result, these juices lead to the intake of large amounts of calories without providing nutrients for the body. In addition, performing chemical tests to determine the quality of fruit juices are time-consuming, destructive, and costly and requires several glassware and reagents. E-nose could be used for real-time monitoring of the volatile components of the food to evaluate different features of the product. Generally, E-nose evaluates a mixture of smells released from a sample and is a reliable, nondestructive, cost-effective, and portable method with high feasibility and speed as well as simple use.

VFF in MMU: Determination for Velocity of Fluid Flow in Multiphase Measurement Unit - Comparative Analysis

The velocity measurement of liquid flow in a channel is a challenging task still. The chemical reaction and heat transfer condition are one of the internal elements of liquid for any process and production industries. Besides, the flow velocity is a significant factor to measure temperature in liquid flow. This research article reviews an overview of the velocity of fluid measurement techniques by advanced concepts in the multiphase measurement system. Based on luminescence properties, the velocity measurement is derived by large particle spectrum analysis due to laser excitation in the machine. This mathematical model analysis is used to measure the velocity of a fluid with the same particles of velocimetry. This development can be derived from the many changes of measurement factors in the heat transfer mechanism. The flowmeter design will be optimized with this mathematical proof for phosphor thermometry measurement technique. This research article contains phosphor thermometry for the measurement with implementing techniques and how this thermometry will be appropriated for temperature measurement in liquid flow. It gives the compare graphic representation for various work of temperature measurement in liquid flow of common aspects. The successful metric measurement can be ended by various intrinsic keys to the future development of the procedure. The velocity measurement performs by LDA and PIV methods. The advantages and limitations have been discussed for both the method at most recent.

A comparison study of polynomial-based PCA, KPCA, LDA and GDA feature extraction methods for epileptic and eye states EEG signals detection using kernel machines

Background and objectiveInitially, analysis of Electroencephalogram (EEG) signals was purely visual, tedious, time-consuming, and required a physician. Changing this old approach to classification proves to be an extraordinary task that gained much attention and a great deal of effort. With this intention, this comparison study focused on the development of polynomial-based feature extraction methods for epileptic and eye states EEG signals detection using kernel machines. MethodPolynomial transforms are applied to decompose EEG signals in the frequency domain before their analysis using linear and non-linear measures. Thereafter, the standard and kernel extension methods are applied to determine principal components and discriminants which help to extract informative and discriminative low-dimensional features. For direct detection of EEG signals, extracted features are fed into kernel machines namely simple multilayer perceptron neural network (sMLPNN) and least-square support vector machine (LS-SVM). ResultsUsing the publicly available Bonn-University database, experimental results demonstrated that features extracted using kernel methods are more discriminative than the ones using standard methods. In addition, compared to the LS-SVM, polynomial-based features with sMLPNN gained higher performances. Moreover, obtained predictivity, accuracy, and area under receiver operating curve also demonstrate that kernel machines can detect epileptic and eye states EEG signals with highest performances of 100%, 100% and 1, respectively. ConclusionThus, the proposed framework can be efficient for EEG diagnosis. Overall, given the complexity and heterogeneity of the brain, it is likely frameworks of this type that will be required to configure intelligent devices for treating epilepsy and to configure eye-brain-computer interface.