Text Data Processing Research Articles

ANALISIS SENTIMEN MASYARAKAT TERHADAP VAKSINASI COVID-19 PADA MEDIA SOSIAL TWITTER MENGGUNAKAN ALGORITMA SUPPORT VECTOR MACHINE (SVM)

This research was conducted to find information about the tendency of Indonesian people regarding the Covid-19 vaccination. The method that the author uses is by collecting data from Twitter social media using the API key provided by Twitter. The process of collecting data using a Python application with several libraries such as tweepy, pandas, numpy and nltk. After the data is crawled, then the data is cleaned with several data cleaning processes such as remove username, remove url, lower case, remove stopwords and lemmatize. Then the results are labeled with the textblob and sklearn libraries. then the data is analyzed using the Support Vector Machine (SVM) algorithm with the best comparison being 20 testing data and 80 training data or as many as 942 testing data and 3766 training data, the prediction results for testing data are f1 score 0.93, accuracy score 0.88, precision score 0.88 and a recall score of 0.99. The results showed that from 4,078 tweet data, there were 2,525 positive sentiments (43.0%), 771 negative sentiments (16.4%), and 1,912 neutral sentiments (40.6%). The results of 80% (3766) of training data and 20% (942) of test data obtained an accuracy score of 73.6%. From this study, it can be concluded that the tendency of Indonesian people when sampling data is taken is more accepting (positive responses) to government policies regarding the Covid-19 vaccination program. In the future, it is hoped that there will be a library that supports text data processing such as regional languages, because researchers found that during the data cleaning process there was a lot of word elimination, because many regional languages ​​were used by the Indonesian people in writing on social media.

Text Data Processing and Classification Algorithm Based on Data Fusion and Granular Computing

With the rapid expansion of network information and the emergence of a large number of electronic texts, how to organize and manage this massive information has become a major challenge. Automatic text categorization technology is to study how to let the machine classify unknown text through self-learning, thus solving the difficulties encountered in manual classification. Because granular computing can reduce the knowledge in solving complex problems, it is more convenient to summarize and acquire knowledge. It has become a hotspot in recent years, and it also provides new ideas for text classification research. The rough set model of granular computing can acquire knowledge by mining decision rules. The decision process is more transparent and easy to understand. It has been paid attention to and applied in text classification research. Based on the research of existing achievements, this paper makes a further study on the application of granular computing in text categorization. After analyzing the existing feature selection methods, the feature distribution is proposed based on the relationship between feature words and categories. By calculating the distribution distance between any two feature words, the feature words with similar distribution distances are aggregated, which effectively reduces the dimension of the feature space and also avoids the individual samples caused by the existing feature selection algorithm. A phenomenon that is discarded due to features. The experimental results show that the clustering method can obtain higher classification accuracy than other feature selection methods when using SVM as the classifier. SVM performs best, and the final text classification accuracy rate can reach 85.46%. According to the correlation principle of the rough set, feature selection is made for each information granularity, the selected feature is used as the condition attribute and the coordination matrix is constructed, and the most similar sample is heuristically searched to obtain the attribute reduction set.

Унифицированное применение методов онтологического инжиниринга в задачах обработки текстовых данных

Unified application issues of ontological engineering methods and tools for text data processing automation in 2 problems classes solving are considered. The first class is the relational database applications construction automation based on the necessary facts extracting directly from unstructured text documents collections in natural language. The second class is the text data processing automation in artificial languages in the program code verifying process for database applications. The information systems designing and developing problems, same as program code verifying problems, are complex for non-professional programmers. The text data processing problem solving automation is also relevant for professional programmers, because it significantly reduces the database applications development time. The ontologically controlled solution concept is proposed for these problems. An unified approach for problems solution and the tools demos implementation are described.

Financial Fraud Detection on Social Networks Based on a Data Mining Approach

The article summarizes the arguments and counter-arguments within the scientific debate on the issue of researching financial frauds in the Internet. The main goal of the research is to develop methodological principles for identifying financial cyber fraud in social networks based on the analysis of comments to identify relevant text patterns that may indicate manipulation attempts and further fraud. The urgency of solving this scientific problem is due to the fact that the mass involvement of Internet users in social interactions in the virtual environment has contributed to the development of various criminal schemes, as well as personal data that is initially entered during registration and information that is published in social networks can be used by a fraudster to carry out illegal financial transactions. The study of the issue of identifying financial fraud in social networks in the article is carried out in the following logical sequence: collecting comments with a corresponding request under publications in the social network using the Instaloader tool; combining comments into groups based on content similarity; conducting preliminary processing of text data (decomposing the text into simpler components (tokens) and reducing similar word forms to their main dictionary form); determination of the level of similarity of text data using the cosine of similarity; building clusters of text data that can indicate the presence of signs of financial fraud under relevant comments in social networks. Instagram was chosen to identify fraudulent operations in social networks. The analysis of comments on the social network Instagram to identify text patterns showed that offers and appeals from specific groups of people and promoted in comments with the help of spam are dangerous. Based on the results of the study, it was concluded that national regulators need to strengthen public control of the Internet, as well as improve the security system at the technical level by using the latest machine learning methods to identify attempts to commit illegal actions with the subsequent imposition of sanctions on such users in social networks.

TempTracker: A Service Oriented Temporal Natural Language Processing Based Tool for Document Data Characterization and Social Network Analysis

With the advent of Web 2.0 based technology, news sites and micro-blog sites have become popular and have attracted the attention of people around the world. Existing textual data captured by these sites is highly beneficial for extracting (a) new information to analyze, and (b) temporal course of change in entities, topics and sentiment for differing granularities. This has been demonstrated by the study described in this paper. After collecting the data, several directions have been investigated in order to demonstrate its effectiveness under the umbrella of entity extraction, topic and sentiment analysis using Natural Language Processing (NLP) tools, temporal social media analysis, and time varying trend results of entity and sentiment aspect of entities. A service-based architecture has been proposed to process text data with NLP tools and to enrich the data. Text data is collected and processed via NLP tools. It is retrieved upon request for data analysis. The reported results illustrate the applicability and effectiveness of the conducted study.

New Language Models for Spelling Correction

Correcting spelling errors based on the context is a fairly significant problem in Natural Language Processing (NLP) applications. The majority of the work carried out to introduce the context into the process of spelling correction uses the n-gram language models. However, these models fail in several cases to give adequate probabilities for the suggested solutions of a misspelled word in a given context. To resolve this issue, we propose two new language models inspired by stochastic language models combined with edit distance. A first phase consists in finding the words of the lexicon orthographically close to the erroneous word and a second phase consists in ranking and limiting these suggestions. We have applied the new approach to Arabic language taking into account its specificity of having strong contextual connections between distant words in a sentence. To evaluate our approach, we have developed textual data processing applications, namely the extraction of distant transition dictionaries. The correction accuracy obtained exceeds 98% for the first 10 suggestions. Our approach has the advantage of simplifying the parameters to be estimated with a higher correction accuracy compared to n-gram language models. Hence the need to use such an approach.

Text Complexity Classification Data Mining Model Based on Dynamic Quantitative Relationship between Modality and English Context

With the rapid development of mobile internet technology, there are a large number of unstructured data in dynamic data, such as text data, multimedia data, etc., so it is essential to analyze and process these unstructured data to obtain potentially valuable information. This article first starts with the theoretical research of text complexity analysis and analyzes the source of text complexity and its five characteristics of dynamic, complexity, concealment, sentiment, and ambiguity, combined with the expression of user needs in the network environment. Secondly, based on the specific process of text mining, namely, data collection, data processing, and data visualization, it is proposed to subdivide the user demand analysis into three stages of text complexity acquisition, recognition, and expression, to obtain a text complexity analysis based on text mining technology. After that, based on computational linguistics and mathematical-statistical analysis, combined with machine learning and information retrieval technology, the text in any format is converted into a content format that can be used for machine learning, and patterns or knowledge are derived from this content format. Then, through the comparison and research of text mining technology, combined with the text complexity analysis hierarchical structure model, a quantitative relationship complexity analysis framework based on text mining technology is proposed, which is embodied in the use of web crawler technology. Experimental results show that the collected quantitative relationship information is identified and expressed in order to realize the conversion of quantitative relationship information into product features. The market data and text data can be integrated to help improve the model performance and the use of text data can further improve predictions for accuracy.

Penggunaan Karakter Kontrol ASCII Untuk Integrasi Data Pada Hasil Enkripsi Algoritma Caesar Cipher


 
 
 Cryptography is a technique or method for securing data from other unauthorized parties. The substitution algorithm is the simplest algorithm and is classified as a classic in the field of cryptography, for example the Caesar cipher algorithm. ASCII code (American Standard Code for Information Interchange) is a code that contains characters that can be processed by a computer. By type, not all ASCII characters can be printed by the computer. Some characters are not printed or illegible as usual. These unreadable characters are called control characters. The control character can be used to improve the performance of the Caesar Cipher algorithm, because it focuses on processing text data. The application of control characters uses a simple method but has a complex solution. The results of this study are in the form of a new method with the Caesar cipher algorithm as a classical cryptographic method or technique and ASCII characters as the basis for the development of the ciphertext performance resulting from the encryption process. In testing this method, the success rate reaches 100% in securing the contents of the document with a sample of 500 letters. While the possibility of solving the ciphertext results is classified as difficult because the control characters of the ciphertext results that are illegible will make the decryption result multiple interpretations.
 
 

VisQuelle : Visual Question based Elementary Learning Companion a system to Facilitate Learning Word Object Associations

We present an automated, visual question answering based companion – VisQuelle - to facilitate elementary learning of word-object associations. In particular, we attempt to harness the power of machine learning models for object recognition and the understanding of combined processing of images and text data from visual-question answering to provide variety and nuance in the images associated with letters or words presented to the elementary learner. We incorporate elements such as gamification to motivate the learner by recording scores, errors, etc., to track the learner’s progress. Translation is also provided to reinforce word-object associations in the user’s native tongue, if the learner is using VisQuelle to learn a second language.

Transfer learning based recurrent neural network algorithm for linguistic analysis

AbstractEach language is a system of understanding and skills that allow language users to work together, hypothesize, express thoughts, opinions; wishes need to be articulated. Linguistics is the research of these structures in all respects: the composition, usage, and sociology of language, in particular, are the core of linguistics. Machine learning is the research area that allows machines to learn without being specifically scheduled. In linguistics, the design of writing is understood to be a foundation for many distinct company apps and probably the most useful if incorporated with machine learning methods. Research shows that besides text tagging and algorithm training, there are major problems in the field of big data. This article provides a collaborative effort (transfer learning integrated into recurrent neural network) to analyze the distinct kinds of writing between the language's linear and noncomputational sides and to enhance granularity. In addition to this, this article creates a recurrent neural network model for learning and processing text data automatically. RNN based linguistic process is fast and cost‐effective. It analyses the text data automatically and iteratively. This is the main reason for using RNN for the linguistic process. The outcome demonstrates stronger incorporation of granularity into the language from both sides. Comparative results of machine learning algorithms are used to determine the best way to analyze and interpret the structure of the language.

Leveraging Unsupervised Machine Learning to Discover Patterns in Linguistic Health Summaries for Eldercare.

The Center for Eldercare and Rehabilitation Technology, at University of Missouri, has researched the use of smart, unobtrusive sensors for older adult residents' health monitoring and alerting in aging-in-place communities for many years. Sensors placed in the apartments of older adult residents generate a deluge of daily data that is automatically aggregated, analyzed, and summarized to aid in health awareness, clinical care, and research for healthy aging. When anomalies or concerning trends are detected within the data, the sensor information is converted into linguistic health messages using fuzzy computational techniques, so as to make it understandable to the clinicians. Sensor data are analyzed at the individual level, therefore, through this study we aim to discover various combinations of patterns of anomalies happening together and recurrently in the older adult's population using these text summaries. Leveraging various computational text data processing techniques, we are able to extract relevant analytical features from the health messages. These features are transformed into a transactional encoding, then processed with frequent pattern mining techniques for association rule discovery. At individual level analysis, resident ID 3027 was considered as an exemplar to describe the analysis. Seven combinations of anomalies/rules/associations were discovered in this resident, out of which rule group three showed an increased recurrence during the COVID lockdown of facility. At the population level, a total of 38 associations were discovered that highlight the health patterns, and we continue to explore the health conditions associated with them. Ultimately, our goal is to correlate the combinations of anomalies with certain health conditions, which can then be leveraged for predictive analytics and preventative care. This will improve the current clinical care systems for older adult residents in smart sensor, aging-in-place communities.

Diagnosing post-traumatic stress disorder using electronic medical record data.

This study proposes a predictive model that uses structured data and unstructured narrative notes from Electronic Medical Records to accurately identify patients diagnosed with Post-Traumatic Stress Disorder (PTSD). We utilize data from primary care clinicians participating in the Manitoba Primary Care Research Network (MaPCReN) representing 154,118 patients. A reference sample of 195 patients that had their PTSD diagnosis confirmed using a manual chart review of structured data and narrative notes, and PTSD negative patients is used as the gold standard data for model training, validation and testing. We assess structured and unstructured data from eight tables in the MaPCReN namely, patient demographics, disease case, examinations, medication, billing records, health condition, risk factors, and encounter notes. Feature engineering is applied to convert data into proper representation for predictive modeling. We explore serial and parallel mixed data models that are trained on both structured and unstructured data to identify PTSD. Model performances were calculated based on a highly skewed hold-out test dataset. The serial model that uses both structured and text data as input, yielded the highest values in sensitivity (0.77), F-measure (0.76), and AUC (0.88) and the parallel model that uses both structured and text data as the input obtained the highest positive predicted value (PPV) (0.75). Diseases such as PTSD are difficult to diagnose. Information recorded in the chart note over multiple visits of the patients with the primary care physicians has higher predictive power than structured data and combining these two data types can increase the predictive capabilities of machine learning models in diagnosing PTSD. While the deep-learning model outperformed the traditional ensemble model in processing text data, the ensemble classifier obtained better results in ingesting a combination of features obtained from both data types in the serial mixed model. The study demonstrated that unstructured encounter notes enhance a model's ability to identify patients diagnosed with PTSD. These findings can enhance quality improvement, research, and disease surveillance related to PTSD in primary care populations.

An overview of complex data stream ensemble classification

With the rapid development of information technology, data streams in various fields are showing the characteristics of rapid arrival, complex structure and timely processing. Complex types of data streams make the classification performance worse. However, ensemble classification has become one of the main methods of processing data streams. Ensemble classification performance is better than traditional single classifiers. This article introduces the ensemble classification algorithms of complex data streams for the first time. Then overview analyzes the advantages and disadvantages of these algorithms for steady-state, concept drift, imbalanced, multi-label and multi-instance data streams. At the same time, the application fields of data streams are also introduced which summarizes the ensemble algorithms processing text, graph and big data streams. Moreover, it comprehensively summarizes the verification technology, evaluation indicators and open source platforms of complex data streams mining algorithms. Finally, the challenges and future research directions of ensemble learning algorithms dealing with uncertain, multi-type, delayed, multi-type concept drift data streams are given.

APPLICATION OF THE PROFESSIONALLY ORIENTED TASKS IN LEARNING INFORMATION TECHNOLOGIES AT HIGHER EDUCATION ESTABLISHMENTS

The most often used information technologies in the future professional activity of students of pedagogical universities are computer technologies of text, tabular and graphic data processing, work with multimedia data, as well as computer telecommunications. Mastering the skills of working with these technologies involves the study of office suites that allow you to perform data processing work. The use of information and communication technologies provides ample opportunities for the organization of educational and cognitive activities of students. Therefore, it is important to teach students the effective use of information and communication technologies in their professional activities. As far as much attention is paid not only to the knowledge that graduates must master, but also the formation of their professional skills, the most modern requirements for higher education meets a new type of educational tasks – professionally-oriented tasks.
 The article clarifies the main aspects of the introduction of professionally-oriented tasks in the educational process of pedagogical higher educational institutions. New possibilities and ways of using a package of office programs in the work of a teacher have been investigated. The professionally-oriented tasks have been applied in creating an educational and methodological complex for students in the study of Microsoft Office 2010. An example of step-by-step and sequential system of tasks for obtaining a certain outcome result taking into account the professional orientation of students has been given. New approaches to the introduction of professionally-oriented tasks in teaching the course “Modern information technologies in the educational process” have been developed. They will allow students to apply skillfully the acquired knowledge and skills to work with office suite programs in their studies, flexibly adapt to situations that arise in the professional activities of teachers, independently acquire the necessary knowledge, effectively use them to solve various problems in practice. A set of interrelated methods was used to solve the tasks included in the study: theoretical: analysis and generalization of scientific and pedagogical literature, modeling of the educational process; empirical: observation of the educational process at school during pedagogical practice and at the university. The introduction of information technology into the educational process of higher education allows teachers to qualitatively change the content, methods and organizational forms of learning.

Intelligent processing of unstructured textual data in document based NoSQL databases

Many organizations view themselves as market driven in this digitalized society, where the decision making depends solely on the customer feedbacks and suggestions. This massive amount of unstructured data produced, affects query processing performance adding to the difficulty in managing or retrieving the data. Many organizations are progressively exploring ways to comprehend and analyse this huge unstructured data to infer operational insights from it. We could solve many of the storage issues of unstructured data using the document-based NoSQL databases. But applying artificial intelligence to make exact matching predictions with the text data is remaining as a challenge. With the new method proposed here, the intelligent processing of the textual data stored in document-based NoSQL databases is made possible, and it is really an added benefit to those, who opt or use NoSQL databases for unstructured data storage.

Sentiment Analysis Approach for Analyzing iPhone Release using Support Vector Machine

Sentiment analysis is a process of understanding, extracting, and processing textual data automatically to get sentiment information contained in a comment sentence on Twitter. Sentiment analysis needs to be done because the use of social media in society is increasing so that it affects the development of public opinion. Therefore, it can be used to analyze public opinion by applying data science, one of which is Natural Language Processing (NLP) and Text Mining or also known as text analytics. The stages of the overall method used in this study are to do text mining on the Twitter site regarding iPhone Release with methods of scraping, labeling, preprocessing (case folding, tokenization, filtering), TF-IDF, and classification of sentiments using the Support Vector Machine. The Support Vector Machine is widely used as a baseline in text-related tasks with satisfactory results, on several evaluation matrices such as accuracy, precision, recall, and F1 score yielding 89.21%, 92.43%, 95.53%, and 93.95, respectively.

The quest for better clinical word vectors: Ontology based and lexical vector augmentation versus clinical contextual embeddings

BackgroundWord vectors or word embeddings are n-dimensional representations of words and form the backbone of Natural Language Processing of textual data. This research experiments with algorithms that augment word vectors with lexical constraints that are popular in NLP research and clinical domain constraints derived from the Unified Medical Language System (UMLS). It also compares the performance of the augmented vectors with Bio + Clinical BERT vectors which have been trained and fine-tuned on clinical datasets. MethodsWord2vec vectors are generated for words in a publicly available de-identified Electronic Health Records (EHR) dataset and augmented by ontologies using three algorithms that have fundamentally different approaches to vector augmentation. The augmented vectors are then evaluated alongside publicly available Bio + Clinical BERT on their correlation with human-annotated lists using Spearman's correlation coefficient. They are also evaluated on the downstream task of Named Entity Recognition (NER). Quantitative and empirical evaluations are used to highlight the strengths and weaknesses of the different approaches. ResultsThe counter-fitted word2vec vectors augmented with information from the UMLS ontology produced the best correlation overall with human-annotated evaluation lists (Spearman's correlation of 0.733 with mini mayo-doctors’ annotation) while Bio + Clinical BERT produces the best results in the NER task (F1 of 0.87 and 0.811 on the i2b2 2010 and i2b2 2012 datasets respectively) in our experiments. ConclusionClinically adapted word2vec vectors successfully encapsulate concepts of lexical and clinical synonymy and antonymy and to a smaller extent, hyponymy and hypernymy. Bio + Clinical BERT vectors perform better at NER and avoid out-of-vocabulary words.

АНГЛОМОВНИЙ БАЗИС МОВИ ПРОГРАМУВАННЯ PYTHON

The dynamic nature of the Python programming language, the accumulation of a certain linguosemiotic basis indicates the similarity of this language with the English language, which is the international one and mediates human communication in both real and virtual worlds. In this study, the English language is positioned as the linguistic basis of Python language of programming, which is widely used in industry, research, natural language processing, textual information retrieval, textual data processing, texts corpora, and more. English language, its lexical features, text representation and interaction with logical and functional basis in the context of Python programming language are considered further in this research. Thus, the unity of verbal units and symbols in the modern English-language digital discourse indicates both the order and variability of the constituents therein. The functionality of linguosemiotic elements produces a network of relationships, where each of these integrated elements can produce from a word or symbol a holistic set of units, which are extrapolated in the English-language digital discourse and mediates human communication with a machine. An overview of the basic properties of Python language, such as values, types, expressions, and operations are in focus of the study. Though users understand the responses of Python interpreter, there is a need to follow certain instructions and codes. To facilitate work with this programming language and prescribed English-language commands, it is necessary to involve linguists to cooperate with programmers to invent a certain logical and reasonable principle of Python commands operation.

Automatic analysis of insurance reports through deep neural networks to identify severe claims

AbstractIn this paper, we develop a methodology to automatically classify claims using the information contained in text reports (redacted at their opening). From this automatic analysis, the aim is to predict if a claim is expected to be particularly severe or not. The difficulty is the rarity of such extreme claims in the database, and hence the difficulty, for classical prediction techniques like logistic regression to accurately predict the outcome. Since data is unbalanced (too few observations are associated with a positive label), we propose different rebalance algorithm to deal with this issue. We discuss the use of different embedding methodologies used to process text data, and the role of the architectures of the networks.

Text Classification Using Novel Term Weighting Scheme-Based Improved TF-IDF for Internet Media Reports

With the rapid development of the internet technology, a large amount of internet text data can be obtained. The text classification (TC) technology plays a very important role in processing massive text data, but the accuracy of classification is directly affected by the performance of term weighting in TC. Due to the original design of information retrieval (IR), term frequency-inverse document frequency (TF-IDF) is not effective enough for TC, especially for processing text data with unbalanced distributions in internet media reports. Therefore, the variance between the DF value of a particular term and the average of all DFsDF¯, namely, the document frequency variance (ADF), is proposed to enhance the ability in processing text data with unbalanced distribution. Then, the normal TF-IDF is modified by the proposed ADF for processing unbalanced text collection in four different ways, namely, TF-IADF, TF-IADF+, TF-IADFnorm, and TF-IADF+norm. As a result, an effective model can be established for the TC task of internet media reports. A series of simulations have been carried out to evaluate the performance of the proposed methods. Compared with TF-IDF on state-of-the-art classification algorithms, the effectiveness and feasibility of the proposed methods are confirmed by simulation results.