Text clustering is the task of grouping a set of texts so that text in the same group will be more similar than those from a different group. The process of grouping text manually requires a significant amount of time and labor. Therefore, automation utilizing machine learning is necessary. One of the most frequently used method to represent textual data is Term Frequency Inverse Document Frequency (TFIDF). However, TFIDF cannot consider the position and context of a word in a sentence. Bidirectional Encoder Representation from Transformers (BERT) model can produce text representation that incorporates the position and context of a word in a sentence. This research analyzed the performance of the BERT model as data representation for text. Moreover, various feature extraction and normalization methods are also applied for the data representation of the BERT model. To examine the performances of BERT, we use four clustering algorithms, i.e., k-means clustering, eigenspace-based fuzzy c-means, deep embedded clustering, and improved deep embedded clustering. Our simulations show that BERT outperforms TFIDF method in 28 out of 36 metrics. Furthermore, different feature extraction and normalization produced varied performances. The usage of these feature extraction and normalization must be altered depending on the text clustering algorithm used.

Recently, the popularity of big data as a research field has shown continuous and wide-scale growth. This study aims to capture the scientific structure and topic evolution of big data research using bibliometrics and text mining-based analysis methods. Bibliographic data of journal articles regarding big data published between 2009 to 2018 were collected from the Scopus database and analyzed. The results show a significant growth of publications since 2014. Furthermore, the findings of this study highlight the core journals, most cited articles, top productive authors, countries, and institutions. Secondly, a unique approach to identifying and analyzing major research themes in big data publications was proposed. Keywords were clustered, and each cluster was labeled as a theme. Moreover, the papers were divided into four sub-periods to observe the thematic evolution. The theme mapping reveals that research on big data is dominated by big data analytics, which covers methods, tools, supporting infrastructure, and applications. Other critical aspects of big data research are security and privacy. Social networks and the Internet of things are significant sources of big data, and the resources and services offered by cloud computing strongly support the management and processing of big data.

The photoplethysmography (PPG) method for continuous noninvasive measurements of blood pressure (BP) offers a more comfortable solution than conventional methods. The main challenge in using the PPG method is that its accuracy is greatly influenced by motion artifacts. In addition, the characteristics of PPG vary depending on physiological conditions; hence, the system must be calibrated to adjust for such changes. We attempt to address these limitations and propose a novel method for the classification of BP using a bidirectional long short-term memory (BLSTM) network with time-frequency (TF) analysis based on PPG signals. The TF analysis extracts information from PPG signals using a short-time Fourier transform (STFT) in the time domain to produce two features, namely, the instantaneous frequency and spectral entropy. Training the BLSTM network using TF features significantly improves the classification performance and decreases the training time. We classify 900 PPG waveform segment samples from 219 adult subjects into three classification levels: normotension (NT), prehypertension (PHT) and hypertension (HT). The results show that the proposed method is successful in the classification of BP with accuracy, sensitivity, and specificity values of 97.33%, 100%, and 94.87%, respectively. The F1 scores of three BP classifications were 97.29%, 97.39%, and 93.93%, respectively. A comparison of current and previous approaches to the classification of BP is accomplished. Our proposed method achieves a higher accuracy than convolutional neural networks (CNNs), k-nearest neighbors (KNN), bagged tree, logistic regression, and AdaBoost tree methods.

Monitoring systems for the early detection of diabetes are essential to avoid potential expensive medical costs. Currently, only invasive monitoring methods are commercially available. These methods have significant disadvantages as patients experience discomfort while obtaining blood samples. A non-invasive method of blood glucose level (BGL) monitoring that is painless and low-cost would address the limitations of invasive techniques. Photoplethysmography (PPG) collects a signal from a finger sensor using a photodiode, and a nearby infrared LED light. The combination of the PPG electronic circuit with artificial intelligence makes it possible to implement the classification of BGL. However, one major constraint of deep learning is the long training phase. We try to overcome this limitation and offer a concept for classifying type 2 diabetes (T2D) using a machine learning algorithm based on PPG. We gathered 400 raw datasets of BGL measured with PPG and divided these points into two classification levels, according to the National Institute for Clinical Excellence, namely, “normal” and “diabetes”. Based on the results for testing between the models, the ensemble bagged trees algorithm achieved the best results with an accuracy of 98%.