Linear SVC Research Articles

Automated Categorization of Research Papers with MONO Supervised Term Weighting in RECApp

Natural Language Processing, specifically text classification or text categorization, has become a trend in computer science. Commonly, text classification is used to categorize large amounts of data to allocate less time to retrieve information. Students, as well as research advisers and panelists, take extra effort and time in classifying research documents. To solve this problem, the researchers used state-of-the-art supervised term weighting schemes, namely: TF-MONO and SQRTF-MONO and its application to machine learning algorithms: K-Nearest Neighbor, Linear Support Vector, Naive Bayes Classifiers, creating a total of six classifier models to ascertain which of them performs optimally in classifying research documents while utilizing Optical Character Recognition for text extraction. The results showed that among all classification models trained, SQRTF-MONO and Linear SVC outperformed all other models with an F1 score of 0.94 both in the abstract and the background of the study datasets. In conclusion, the developed classification model and application prototype can be a tool to help researchers, advisers, and panelists to lessen the time spent in classifying research documents.

Comparison of Support Vector Machines and K-Nearest Neighbor Algorithm Analysis of Spam Comments on Youtube Covid Omicron

Every time a new variant of Coronavirus (Covid-19) appears, themedia or news platforms review it to find out whether the new variantis more dangerous or contagious than before. One of the media orplatforms that is fast in presenting news in videos is YouTube.YouTube is a social media that can upload videos, watch videos, andcomment on the video. The comment field on YouTube videos cannotbe separated from spam comments that annoy other users who want tofollow or participate in the comment column. Indication of spamcomments is still done by observing one by one; this is very inefficientand time-consuming. This study aims to create a model that canclassify spam on YouTube comments. The classification method uses the SVM (Support Vector Machines) algorithm and the KNN (K-Nearest Neighbor) algorithm to identify spam comments or not with comment data taken from Omicron's Covid-19 news video on national news channels. The classification results show that the SVM method is superior inaccuracy with the Linear SVC algorithm of 75.12%, SVC of 76.11%, and Nu-SVC of 77.11%. While the KNN algorithm with k=2 is 65.67%, k=4 is 64.51%, k=6 is 62.35%.

Multi-label multi-class COVID-19 Arabic Twitter dataset with fine-grained misinformation and situational information annotations.

Since the inception of the current COVID-19 pandemic, related misleading information has spread at a remarkable rate on social media, leading to serious implications for individuals and societies. Although COVID-19 looks to be ending for most places after the sharp shock of Omicron, severe new variants can emerge and cause new waves, especially if the variants can evade the insufficient immunity provided by prior infection and incomplete vaccination. Fighting the fake news that promotes vaccine hesitancy, for instance, is crucial for the success of the global vaccination programs and thus achieving herd immunity. To combat the proliferation of COVID-19-related misinformation, considerable research efforts have been and are still being dedicated to building and sharing COVID-19 misinformation detection datasets and models for Arabic and other languages. However, most of these datasets provide binary (true/false) misinformation classifications. Besides, the few studies that support multi-class misinformation classification deal with a small set of misinformation classes or mix them with situational information classes. False news stories about COVID-19 are not equal; some tend to have more sinister effects than others (e.g., fake cures and false vaccine info). This suggests that identifying the sub-type of misinformation is critical for choosing the suitable action based on their level of seriousness, ranging from assigning warning labels to the susceptible post to removing the misleading post instantly. We develop comprehensive annotation guidelines in this work that define 19 fine-grained misinformation classes. Then, we release the first Arabic COVID-19-related misinformation dataset comprising about 6.7K tweets with multi-class and multi-label misinformation annotations. In addition, we release a version of the dataset to be the first Twitter Arabic dataset annotated exclusively with six different situational information classes. Identifying situational information (e.g., caution, help-seeking) helps authorities or individuals understand the situation during emergencies. To confirm the validity of the collected data, we define three classification tasks and experiment with various machine learning and transformer-based classifiers to offer baseline results for future research. The experimental results indicate the quality and validity of the data and its suitability for constructing misinformation and situational information classification models. The results also demonstrate the superiority of AraBERT-COV19, a transformer-based model pretrained on COVID-19-related tweets, with micro-averaged F-scores of 81.6% and 78.8% for the multi-class misinformation and situational information classification tasks, respectively. Label Powerset with linear SVC achieved the best performance among the presented methods for multi-label misinformation classification with micro-averaged F-scores of 76.69%.

A retrospective analysis based on multiple machine learning models to predict lymph node metastasis in early gastric cancer.

Endoscopic submucosal dissection has become the primary option of treatment for early gastric cancer. However, lymph node metastasis may lead to poor prognosis. We analyzed factors related to lymph node metastasis in EGC patients, and we developed a construction prediction model with machine learning using data from a retrospective series. Two independent cohorts' series were evaluated including 305 patients with EGC from China as cohort I and 35 patients from Spain as cohort II. Five classifiers obtained from machine learning were selected to establish a robust prediction model for lymph node metastasis in EGC. The clinical variables such as invasion depth, histologic type, ulceration, tumor location, tumor size, Lauren classification, and age were selected to establish the five prediction models: linear support vector classifier (Linear SVC), logistic regression model, extreme gradient boosting model (XGBoost), light gradient boosting machine model (LightGBM), and Gaussian process classification model. Interestingly, all prediction models of cohort I showed accuracy between 70 and 81%. Furthermore, the prediction models of the cohort II exhibited accuracy between 48 and 82%. The areas under curve (AUC) of the five models between cohort I and cohort II were between 0.736 and 0.830. Our results support that the machine learning method could be used to predict lymph node metastasis in early gastric cancer and perhaps provide another evaluation method to choose the suited treatment for patients.

Hate and Aggression Analysis in NLP with Explainable AI

Social platforms such as Twitter and Facebook have now become only media to express their thoughts, and due to lack of censorship, it often embellishes themselves as an abode for hate towards minorities. People of color, Asian people, Muslims, women, transgenders, and LGBTQ+ communities are often the target of such online hate and aggression. Though several companies have incorporated considerable algorithms on their platforms, nevertheless due to being rather hard to often detect such comments still make it to the platforms, creating a negative space towards targeted people. This research involves the study and comparison of different hate and aggression detection algorithms with intent on two languages, i.e. English and German including machine learning models (linear SVC, logistic regression, multinomial naive Bayes and random forests) with their variations with feature engineering and bag of words and deep learning (CNN-GRU static, TCN static, Seq2Seq) with their variations vis-à-vis Word2Vec embedding. CNN+GRU static + Word2Vec embedding has outperformed all the other techniques with an accuracy of 68.29%.

Classifying Illegal Activities on Tor Network using Hybrid Technique

With the freedom offered by the Deep Web, people have the opportunity to express themselves freely and discretely, and sadly, this is one of the reasons why people carry out illicit activities there. In this work, a novel dataset for Dark Web active domains known as crawler-DB is presented. To build the crawler-DB, the Onion Routing Network (Tor) was sampled, and then a web crawler capable of crawling into links was built. The link addresses that are gathered by the crawler are then classified automatically into five classes. The algorithm built in this study demonstrated good performance as it achieved an accuracy of 85%. A popular text representation method was used with the proposed crawler-DB crossed by two different supervised classifiers to facilitate the categorization of the Tor concealed services. The results of the experiments conducted in this study show that using the Term Frequency-Inverse Document Frequency (TF-IDF) word representation with a linear support vector classifier achieves 91% of 5 folds cross-validation accuracy when classifying a subset of illegal activities from crawler-DB, while the accuracy of Naïve Bayes was 80.6%. The good performance of the linear SVC might support potential tools to help the authorities in the detection of these activities. Moreover, outcomes are expected to be significant in both practical and theoretical aspects, and they may pave the way for further research.

Extraction of Meaningful Information from Unstructured Clinical Notes Using Web Scraping

In the medical field, the clinical notes taken by the doctor, nurse, or medical practitioner are considered to be one of the most important medical documents. These documents hold information regarding the patient including the patient’s current condition, family history, disease, symptoms, medications, lab test reports, and other vital information. Despite these documents holding important information regarding the patients, they cannot be used as the data are unstructured. Organizing a huge amount of data without any mistakes is highly impossible for humans, so ignoring unstructured data is not advisable. Hence, to overcome this issue, the web scraping method is used to extract the clinical notes from the Medical Transcription (MT) samples which hold many transcripted clinical notes of various departments. In the proposed method, Natural Language Processing (NLP) is used to pre-process the data, and the variants of the Term Frequency-Inverse Document Frequency (TF-IDF)-based vector model are used for the feature selection, thus extracting the required data from the clinical notes. The performance measures including the accuracy, precision, recall and F1 score are used in the identification of disease, and the result obtained from the proposed system is compared with the best performing machine learning algorithms including the Logistic Regression, Multinomial Naive Bayes, Random Forest classifier and Linear SVC. The result obtained proves that the Random Forest Classifier obtained a higher accuracy of 90% when compared to the other algorithms.

REVIEW SPAM DETECTION USING MACHINE LEARNING

Reviews, ratings, and experience stories left by customers on e-commerce websites and other online services are helpful to both buyers and sellers. The reviewer can foster more brand loyalty and aid in the understanding of other consumers' product experiences. Similar to how reviews help customers earn more profiles, reviews help businesses sell more things by enhancing customer satisfaction. However, suppliers may, unfortunately, abuse these review processes. For instance, one might fabricate positive evaluations to boost the reputation of a brand or attempt to denigrate rival brands' goods by posting fictitious reviews about them. Utilizing various machine learning techniques and tools are examples of existing solutions with supervision. Unlike previous work, I decided to use a wide range of vocabulary to work with, such as many datasets integrated into one enormous data set, rather than a confined dataset. Based on the reviews' text and emoji usage, sentiment analysis has been implemented. Review fraud is recognised and classed. The use of the Linear SVC, Support Vector Machine, and Random Forest algorithms yields the test results.

Identification of high-risk intracranial plaques with 3D high-resolution magnetic resonance imaging-based radiomics and machine learning.

Identifying high-risk intracranial plaques is significant for the treatment and prevention of stroke. To develop a high-risk plaque model using three-dimensional (3D) high-resolution magnetic resonance imaging (HRMRI) based radiomics features and machine learning. 136 patients with documented symptomatic intracranial artery stenosis and available HRMRI data were included. Among these patients, 136 and 92 plaques were identified as symptomatic and asymptomatic plaques, respectively. A conventional model was developed by recording and quantifying the radiological plaque characteristics. Radiomics features from T1-weighted images (T1WI) and contrast-enhanced T1WI (CE-T1WI) were used to construct a high-risk plaque model with linear support vector classification (linear SVC). The radiological and radiomics features were combined to build a combined model. Receiver operating characteristic (ROC) curves were used to evaluate these models. Plaque length, burden, and enhancement were independently associated with clinical symptoms and were included in the conventional model, which had an AUC of 0.853 vs. 0.837 in the training and test sets. While the radiomics and the combined model showed an improved AUC: 0.923 vs. 0.925 for the training sets and 0.906 vs. 0.903 in the test sets. Both the radiomics model (p = 0.024, p = 0.018) and combined model (p = 0.042, p = 0.049) outperformed the conventional model in the two sets, whereas the performance of the combined model was not significantly different from that of the radiomics model in the two sets (p = 0.583 and p = 0.606). The radiomics model based on 3D HRMRI can accurately differentiate symptomatic from asymptomatic intracranial arterial plaques and significantly outperforms the conventional model.

Effect of public sentiment on stock market movement prediction during the COVID-19 outbreak.

Forecasting the stock market is one of the most difficult undertakings in the financial industry due to its complex, volatile, noisy, and nonparametric character. However, as computer science advances, an intelligent model can help investors and analysts minimize investment risk. Public opinion on social media and other online portals is an important factor in stock market predictions. The COVID-19 pandemic stimulates online activities since individuals are compelled to remain at home, bringing about a massive quantity of public opinion and emotion. This research focuses on stock market movement prediction with public sentiments using the long short-term memory network (LSTM) during the COVID-19 flare-up. Here, seven different sentiment analysis tools, VADER, logistic regression, Loughran–McDonald, Henry, TextBlob, Linear SVC, and Stanford, are used for sentiment analysis on web scraped data from four online sources: stock-related articles headlines, tweets, financial news from "Economic Times" and Facebook comments. Predictions are made utilizing both feeling scores and authentic stock information for every one of the 28 opinion measures processed. An accuracy of 98.11% is achieved by using linear SVC to calculate sentiment ratings from Facebook comments. Thereafter, the four estimated sentiment scores from each of the seven instruments are integrated with stock data in a step-by-step fashion to determine the overall influence on the stock market. When all four sentiment scores are paired with stock data, the forecast accuracy for five out of seven tools is at its most noteworthy, with linear SVC computed scores assisting stock data to arrive at its most elevated accuracy of 98.32%.

Multisource Data Integration and Comparative Analysis of Machine Learning Models for On-Street Parking Prediction

Searching for a free parking space can lead to traffic congestion, increasing fuel consumption, and greenhouse gas pollution in urban areas. With an efficient parking infrastructure, the cities can reduce carbon emissions caused by additional fuel combustion, waiting time, and traffic congestion while looking for a free parking slot. A potential solution to mitigating parking search is the provision of parking-related data and prediction. Previously many external data sources have been considered in prediction models; however, the underlying impact of contextual data points and prediction has not received due attention. In this work, we integrated parking occupancy, pedestrian, weather, and traffic data to analyze the impact of external factors on on-street parking prediction. A comparative analysis of well-known Machine (ML) Learning and Deep Learning (DL) techniques, including Multilayer Perceptron (MLP), Random Forest (RF), Decision Trees (DT), K-Nearest Neighbors (KNN), Gradient Boosting (GA), Adaptive Boosting (AB), and linear SVC for the prediction of OnStreet parking space availability has been conducted. The results show that RF outperformed other techniques evaluated with an average accuracy of 81% and an AUC of 0.18. The comparative analysis shows that less complex algorithms like RF, DT, and KNN outperform complex algorithms like MLP in terms of prediction accuracy. All four data sources have positively impacted the prediction, and the proposed solution can determine the best possible parking slot based on weather conditions, traffic flow, and pedestrian volume. The experiments on live prediction showed an ingest rate of 0.1 and throughput of 0.3 events per second, demonstrating a fast and reliable prediction approach for available slots within a 5–10 min time frame. The study is scalable for larger time frames and faster predictions that can be implemented for IoT-based big data-driven environments for on-street and off-street parking.

Analysis of sentiments on the onset of Covid-19 using Machine Learning Techniques

The novel coronavirus (Covid-19) pandemic has struck the whole world and is one of the most striking topics on social media platforms. Sentiment outbreak on social media enduring various thoughts, opinions, and emotions about the Covid-19 disease, expressing views they are feeling presently. Analyzing sentiments helps to yield better results. Gathering data from different blogging sites like Facebook, Twitter, Weibo, YouTube, Instagram, etc., and Twitter is the largest repository. Videos, text, and audio were also collected from repositories. Sentiment analysis uses opinion mining to acquire the sentiments of its users and categorizes them accordingly as positive, negative, and neutral. Analytical and machine learning classification is implemented to 3586 tweets collected in different time frames. In this paper, sentiment analysis was performed on tweets accumulated during the Covid-19 pandemic, Coronavirus disease. Tweets are collected from the Twitter database using Hydrator a web-based application. Data-preprocessing removes all the noise, outliers from the raw data. With Natural Language Toolkit (NLTK), text classification for sentiment analysis and calculate the score subjective polarity, counts, and sentiment distribution. N-gram is used in textual mining -and Natural Language Processing for a continuous sequence of words in a text or document applying uni-gram, bi-gram, and tri-gram for statistical computation. Term frequency and Inverse document frequency (TF-IDF) is a feature extraction technique that converts textual data into numeric form. Vectorize data feed to our model to obtain insights from linguistic data. Linear SVC, MultinomialNB, GBM, and Random Forest classifier with Tfidf classification model applied to our proposed model. Linear Support Vector classification performs better than the other two classifiers. Results depict that RF performs better.

Modelling sentiments based on objectivity and subjectivity with self-attention mechanisms.

Background: The proliferation of digital commerce has allowed merchants to reach out to a wider customer base, prompting a study of customer reviews to gauge service and product quality through sentiment analysis. Sentiment analysis can be enhanced through subjectivity and objectivity classification with attention mechanisms. Methods: This research includes input corpora of contrasting levels of subjectivity and objectivity from different databases to perform sentiment analysis on user reviews, incorporating attention mechanisms at the aspect level. Three large corpora are chosen as the subjectivity and objectivity datasets, the Shopee user review dataset (ShopeeRD) for subjectivity, together with the Wikipedia English dataset (Wiki-en) and Internet Movie Database (IMDb) for objectivity. Word embeddings are created using Word2Vec with Skip-Gram. Then, a bidirectional LSTM with an attention layer (LSTM-ATT) imposed on word vectors. The performance of the model is evaluated and benchmarked against classification models of Logistics Regression (LR) and Linear SVC (L-SVC). Three models are trained with subjectivity (70% of ShopeeRD) and the objectivity (Wiki-en) embeddings, with ten-fold cross-validation. Next, the three models are evaluated against two datasets (IMDb and 20% of ShopeeRD). The experiments are based on benchmark comparisons, embedding comparison and model comparison with 70-10-20 train-validation-test splits. Data augmentation using AUG-BERT is performed and selected models incorporating AUG-BERT, are compared. Results: L-SVC scored the highest accuracy with 56.9% for objective embeddings (Wiki-en) while the LSTM-ATT scored 69.0% on subjective embeddings (ShopeeRD). Improved performances were observed with data augmentation using AUG-BERT, where the LSTM-ATT+AUG-BERT model scored the highest accuracy at 60.0% for objective embeddings and 70.0% for subjective embeddings, compared to 57% (objective) and 69% (subjective) for L-SVC+AUG-BERT, and 56% (objective) and 68% (subjective) for L-SVC. Conclusions: Utilizing attention layers with subjectivity and objectivity notions has shown improvement to the accuracy of sentiment analysis models.

Modelling sentiments based on objectivity and subjectivity with self-attention mechanisms

Background : The proliferation of digital commerce has allowed merchants to reach out to a wider customer base, prompting a study of customer reviews to gauge service and product quality through sentiment analysis. Sentiment analysis can be enhanced through subjectivity and objectivity classification with attention mechanisms. Methods: This research includes input corpora of contrasting levels of subjectivity and objectivity from different databases to perform sentiment analysis on user reviews, incorporating attention mechanisms at the aspect level. Three large corpora are chosen as the subjectivity and objectivity datasets, the Shopee user review dataset (ShopeeRD) for subjectivity, together with the Wikipedia English dataset (Wiki-en) and Internet Movie Database (IMDb) for objectivity. Word embeddings are created using Word2Vec with Skip-Gram. Then, a bidirectional LSTM with an attention layer (LSTM-ATT) imposed on word vectors. The performance of the model is evaluated and benchmarked against classification models of Logistics Regression (LR) and Linear SVC (L-SVC). Three models are trained with subjectivity (70% of ShopeeRD) and the objectivity (Wiki-en) embeddings, with ten-fold cross-validation. Next, the three models are evaluated against two datasets (IMDb and 20% of ShopeeRD). The experiments are based on benchmark comparisons, embedding comparison and model comparison with 70-10-20 train-validation-test splits. Data augmentation using AUG-BERT is performed and selected models incorporating AUG-BERT, are compared. Results: L-SVC scored the highest accuracy with 56.9% for objective embeddings (Wiki-en) while the LSTM-ATT scored 69.0% on subjective embeddings (ShopeeRD). Improved performances were observed with data augmentation using AUG-BERT, where the LSTM-ATT+AUG-BERT model scored the highest accuracy at 60.0% for objective embeddings and 70.0% for subjective embeddings, compared to 57% (objective) and 69% (subjective) for L-SVC+AUG-BERT, and 56% (objective) and 68% (subjective) for L-SVC. Conclusions: Utilizing attention layers with subjectivity and objectivity notions has shown improvement to the accuracy of sentiment analysis models.

Machine Learning Techniques for True and Fake Job Posting

According to research, there are around 188 million unemployed people around the globe. We may find many job vacancies on job portals and across the internet to help the job seekers. India alone has more than a hundred job portals. One major issue people face here is that the job seekers are not sure if the employer is real or fake. Most of these portals do not have a system that could check if the employer posting a job is real or fake. Scammers are making use of this opportunity to post fake job offers which might look genuine to the job seekers applying for it. The poor job seekers might lose a large amount of money and time. A best possible solution for this problem would be that the job portal itself being able to identify if the job being posted is real or fake. Paper suggests using a machine learning model to achieve this goal. An idea here is to use natural language processing to understand and analyze the job posting and then making use of a machine learning model to predict if the job posting is real or fake. The first step is to import a dataset which has real life, real, and fake job posting. In this project, Employment Scam Aegean Dataset provided by University of Aegean Laboratory of Information and Communication system security is being used. Linear SVC is being used in this project for predicting real and fake job posting on a job portal.

An E-mail Classification Algorithm based on Stacking Integrated Learning

<p>The text filtering of traditional anti spam system mainly focuses on keyword matching and text fingerprint analysis, which is difficult to accurately identify and classify spam. Therefore, an integrated learning algorithm based on stackin g is proposed in this paper. Firstly, the algorithm takes the manually marked text data of various categories as samples, uses TF-IDF algorithm to train the word vector space model, then selects linear SVC, xgboost and logistic regression algorithm to structure the base classifier, uses random forest algorithm to structure the meta classifier, and combines the stacking ensemble learning algorithm to structure the classification model. It achieves the function of dividing e-mail into five categories: illegal, advertisement, news, bill and recruitment. From the simulation results, the AUC values of the stacking integrated learning classification algorithm for each category are 0.92, 0.95, 1.00, 0.93 and 0.97 respectively, and the AP values are 0.86, 0.88, 1.00, 0.88 and 0.94 respectively, which realizes the high performance and high precision of text classification.</p> <p> </p>

Prediction of Cardiovascular Risk Using Extreme Learning Machine-Tree Classifier on Apache Spark Cluster

Introduction: In the recent scenario, machine learning is considered a prevailing area in the diverse fields of science and technology include image processing, automobiles, banking, finance, etc. The availability of data and adverse improvements over machine learning techniques have become more feasible to understand and to work on various channels of real-time health analytics. Method: In this paper, a health status prediction system is proposed to detect cardiovascular diseases through patients’ tweets. Further, analytics will be carried on a distributed Apache Spark framework to reduce the time taken for both training and testing when compared with regular standalone machines. Results: Performance of the proposed framework with Extreme Learning Machine (ELM) - Tree classifier is evaluated on two different corpora, and which outperforms other classifiers such as Decision Trees, Naïve Bayes, and Linear SVC, DNN, etc. in both accuracy and time. Discussion: Social media streaming data is considered as one of the major sources for data in the proposed system. Based on the model, the attributes of the incoming user tweets are analyzed, and accordingly cardiovascular risk is predicted. Further, current health status is tweeted back as a reply to the respective user along with a copy to the family and caretakers. Conclusion: This proposed work provides the development of the alert-based system for heart status prediction by adding some additional features impacting the accuracy besides reducing the response time by using Big data Apache Spark Distributed Framework.

Malware Classification with Improved Convolutional Neural Network Model

Malware is a threat to people in the cyber world. It steals personal information and harms computer systems. Various developers and information security specialists around the globe continuously work on strategies for detecting malware. From the last few years, machine learning has been investigated by many researchers for malware classification. The existing solutions require more computing resources and are not efficient for datasets with large numbers of samples. Using existing feature extractors for extracting features of images consumes more resources. This paper presents a Convolutional Neural Network model with pre-processing and augmentation techniques for the classification of malware gray-scale images. An investigation is conducted on the Malimg dataset, which contains 9339 gray-scale images. The dataset created from binaries of malware belongs to 25 different families. To create a precise approach and considering the success of deep learning techniques for the classification of raising the volume of newly created malware, we proposed CNN and Hybrid CNN+SVM model. The CNN is used as an automatic feature extractor that uses less resource and time as compared to the existing methods. Proposed CNN model shows (98.03%) accuracy which is better than other existing CNN models namely VGG16 (96.96%), ResNet50 (97.11%) InceptionV3 (97.22%), Xception (97.56%). The execution time of the proposed CNN model is significantly reduced than other existing CNN models. The proposed CNN model is hybridized with a support vector machine. Instead of using Softmax as activation function, SVM performs the task of classifying the malware based on features extracted by the CNN model. The proposed fine-tuned model of CNN produces a well-selected features vector of 256 Neurons with the FC layer, which is input to SVM. Linear SVC kernel transforms the binary SVM classifier into multi-class SVM, which classifies the malware samples using the one-against-one method and delivers the accuracy of 99.59%.

Modelling sentiments based on objectivity and subjectivity with self-attention mechanisms

Background : The proliferation of digital commerce has allowed merchants to reach out to a wider customer base, prompting a study of customer reviews to gauge service and product quality through sentiment analysis. Sentiment analysis can be enhanced through subjectivity and objectivity classification with attention mechanisms. Methods: This research includes input corpora of contrasting levels of subjectivity and objectivity from different databases to perform sentiment analysis on user reviews, incorporating attention mechanisms at the aspect level. Three large corpora are chosen as the subjectivity and objectivity datasets, the Shopee user review dataset (ShopeeRD) for subjectivity, together with the Wikipedia English dataset (Wiki-en) and Internet Movie Database (IMDb) for objectivity. Word embeddings are created using Word2Vec with Skip-Gram. Then, a bidirectional LSTM with an attention layer (LSTM-ATT) imposed on word vectors. The performance of the model is evaluated and benchmarked against classification models of Logistics Regression (LR) and Linear SVC (L-SVC). Three models are trained with subjectivity (70% of ShopeeRD) and the objectivity (Wiki-en) embeddings, with ten-fold cross-validation. Next, the three models are evaluated against two datasets (IMDb and 20% of ShopeeRD). The experiments are based on benchmark comparisons, embedding comparison and model comparison with 70-10-20 train-validation-test splits. Data augmentation using AUG-BERT is performed and selected models incorporating AUG-BERT, are compared. Results: L-SVC scored the highest accuracy with 56.9% for objective embeddings (Wiki-en) while the LSTM-ATT scored 69.0% on subjective embeddings (ShopeeRD). Improved performances were observed with data augmentation using AUG-BERT, where the LSTM-ATT+AUG-BERT model scored the highest accuracy at 60.0% for objective embeddings and 70.0% for subjective embeddings, compared to 57% (objective) and 69% (subjective) for L-SVC+AUG-BERT, and 56% (objective) and 68% (subjective) for L-SVC. Conclusions: Utilizing attention layers with subjectivity and objectivity notions has shown improvement to the accuracy of sentiment analysis models.

Performance evaluation of a state-of-the-art automotive radar and corresponding modeling approaches based on a large labeled dataset

Radar is a key sensor to achieve a reliable environment perception for advanced driver assistance system and automated driving (ADAS/AD) functions. Reducing the development efforts for ADAS functions and eventually enabling AD functions demands the extension of conventional physical test drives with simulations in virtual test environments. In such a virtual test environment, the physical radar unit is replaced by a virtual radar model. Driving datasets, such as the nuScenes dataset, containing large amounts of annotated sensor measurements, help understand sensor capabilities and play an important role in sensor modeling. This article includes a thorough analysis of the radar data available in the nuScenes dataset. Radar properties, such as detection thresholds, and detection probabilities depending on object, environment, and radar parameters, as well as object properties, such as reflection behavior depending on object type, are investigated quantitatively. The overall detection probability of the considered radar (Continental ARS-408-21) was found to be 27.81%. Four radar models on object level with different complexity levels and different parametrisation requirements are presented: a simple RCS-based radar model with an accuracy of 51%, a linear SVC model with an accuracy of 70%, a Random Forest model with an accuracy of 83%, and a Gradient Boost model with an accuracy of 86%. The feature importance analysis of the machine learning algorithms revealed that object class, object size, and object visibility are the most important parameters for the presented radar models. In contrast, daytime and weather conditions seem to have only minor influence on the modeling results.