Description:Field of Invention
The system and technique used in the current invention analyze textual data to identify coronaviruses using classification—based techniques.
The Objectives of this Invention
The main objective of this innovation is to provide an administration system that is capable of identifying the coronavirus. One method to bring some order to the current turmoil could be to use various AI techniques to detect diseases. The main objective of the current finding is to use the framework with much higher accuracy, which incorporates the classification for extra impacts.
Background of the Invention
The coronavirus, or SARS-COV-2, originated in China and rapidly spread throughout the world. The World Health Organization (WHO) designated the resulting disease as COVID-19. COVID-19 was declared a pandemic on March 11, 2020. Among the initial symptoms of VIRUS are fever, cough, fatigue, and myalgia. In more catastrophic cases, one may experience dyspnea, pneumonia, severe acute respiratory disease, cardiac problems, or even pass away. As soon as possible, it is imperative to identify those who are impacted so that contact tracing and quarantine may be implemented to stop the disease from spreading. In response to this VIRUS, governments everywhere have made proclamations promoting social segregation and self-isolation. (WO2021/231044A1) states that a classification-based technique for coronavirus detection has been disclosed. The system's numerous wearable medical sensors communicate with one or more processors (WMSs). The processors are configured to receive questionnaire data from a user interface in addition to physiological data from WMSs. Additionally, the processors are configured to build at least one coronavirus inference model through the training of at least one neural network using questionnaire data and raw physiological data that have been enhanced with synthetic data and grown and pruned.
An additional application type, CN2020/111834010A, offers a COVID-19 false negative detection method based on XGboost and attribute reduction. The processes in this method are as follows: S1: gathering information from COVID-19 case samples, preprocessing, and enhancing the information; S2: reducing attributes, dimensionality, and dividing the sample data into training and test sets; S3, importance screening on VIRUS detecting core indexes using the XGboost tree-lifting extendable system; S4, training the XGboost method assessment technique and building an evaluation system with the training set data; and S5, identifying the case data using the evaluations algorithm. Test kits and techniques for determining the presence of coronavirus polynucleotides in biological samples are covered in (US2020/10689716B1), another method that was developed. The invention's kit (TR2020/06563A2) includes unique primer and probe sets that supply each of these processes, along with antagonist combinations and tailored enzymatic activity.
NLP has seen a great deal of interest recently, especially in the field of text analysis. Categorization is a crucial task in text analysis that can be accomplished by a range of techniques (Wu et al. [2020], Nature, Vol. 579, pp. 265-269). Kumar et al. [2018], International Journal of Information Technology, 12, pp. 1159–1169, performed a SWOT analysis of several controlled and uncontrolled text classification approaches with the goal of retrieving unorganized information. Text categorization serves various purposes, such as sentiment analysis, fraud detection, and spam recognition. Sentiment mining is mostly used in marketing, elections, and businesses. Verma et al. (2019), International Journal of Recent Technology and Engineering (IJRTE), Vol. 8, No. 3, pp. -8338-8341, assessed the viewpoints of Indian government programs using lexicon-based dictionaries. Because machine learning can efficiently treat disorders like diabetes and seizures, it has changed the way that diagnoses are thought of. Bullock and colleagues (2020), Journal of Artificial Intelligence Research, Vol. 69, pp. 807-845, assert that the accurate diagnosis provided by deep learning and machine learning technology can supplant human judgment. The right diagnosis can save radiologists' time and be more cost-effective than conventional COVID-19 exams. CT, or computed tomography, scans and X-rays may be utilized to train an automatic learning model. There are multiple ongoing projects in this context.
As per the work conducted by Cinnasamy et al. (2022), Materials Today: Proceedings, Volume 64, Part 1, Pages 448–451, we have employed Twitter data to give sentiment analysis. Using the Twitter API, our computer first gathers tweets and hashtags pertaining to different kinds of VIRUS vaccines that have been posted on Twitter. After that, the imported Tweets are automatically configured to generate a set of random and ignorant rules. A diagnostic decision-making tool that aids in the analysis of the victims' lung scans can lessen the doctor's commitment. Convolutional Neural Networks (CNN) VGG16 model, among other machine learning approaches, have been developed for the current study by K. S. Prasad et al. (2022), 2022 International Conference on Computer Communication and Informatics (ICCCI), Coimbatore, India, 2022, pp. 1-5).
Summary of the Invention
The primary technological problems that the prior art encountered are addressed and resolved by the current invention. The current invention makes use of machine learning techniques to forecast VIRUS using clinical text data in order to address these problems. This invention includes multiple modules that use machine learning techniques to examine data and forecast VIRUS disorders.
Detailed Description of the Invention
The WHO declared that the coronavirus pandemic posed an immediate threat. Researchers and organizations are freely providing information on this global pandemic. We obtained the information from the publicly available data source GitHub. The majority of the data, which comprises about 24 attributes, is made up of information on 212 people who have coronavirus and other virus features. Permission. notes on health and other topics. Since our job involves text mining, we acquired medical records and findings. Text makes up clinical notes, but labels for the corresponding text comprise typical results. It was established how long each of the 212 documents was. The text needed to be modified in order to apply machine learning techniques because it was disorganized. During this stage, the text is cleaned up by removing unnecessary text using a variety of techniques.
Refinement of the data is increased through lemmatization and capitalization. The removal of stop words, symbols, links, and URLs increases the precision of the classification. Several characteristics from the altered clinical reports have been eliminated based on interpretation and converted to random values. Relevant characteristics are extracted using the TF//IDF method. Bigrams and unigrams were taken out of the word bag after it was also taken into consideration. Forty relevant qualities that are useful for data classification were identified. The characteristic that corresponds to weight is assigned to classification-based algorithms, which are fed the same data.

Based on textual data, a workable model to identify VIRUS disorders can use text and sentiment analysis. An overview of the potential operation of such a model is provided here:
Data collection: Compiling a dataset containing textual information about VIRUS from news stories, academic papers, social media posts, and medical reports is the initial step in the process. A combination of positive and negative cases, including proven COVID-19 cases as well as cases of other respiratory disorders, should be included in this dataset. Preprocessing of the data: To eliminate noise and standardize the format, the gathered textual data must be preprocessed. Preprocessing could include things like deleting stopwords, changing the text to lowercase, removing unnecessary letters or symbols, and tokenizing the content—splitting it up into individual words or tokens. Feature extraction: After the text data has been preprocessed, pertinent features must be taken out. Term frequency-inverse document frequency (TF-IDF), bag-of-words (BoW), and word embeddings (e.g., Word2Vec or GloVe) are common techniques in text analysis. These methods effectively convey the context and semantic meaning of the text's words. Sentiment analysis: Sentiment analysis techniques can be used to ascertain the sentiment conveyed in the text once the features have been retrieved. The goal of sentiment analysis is to categorize a sentiment as neutral, negative, or positive. This stage can be very important in determining the emotional tone that people in the text data are expressing that is related to the sickness, such as relief, dread, or anxiety. Classification model: A classification model can be trained using the sentiment labels and the preprocessed and feature-extracted data. For this, a variety of machine learning methods can be used, including support vector machines (SVM), random forests, decision trees, and more sophisticated deep learning models like transformers or recurrent neural networks (RNNs). Model training and evaluation: The dataset is split into training and testing sets. The classification model learns the correlations and patterns between the sentiment labels and textual attributes by using the training set. The model's performance is then assessed using the testing set by calculating metrics such as F1-score, accuracy, precision, and recall. Prediction: Following training and assessment, the model can be applied to classify previously unreviewed text data and forecast sentiment. The text will be given a sentiment label by the model that will indicate whether or not it is connected to a COVID-19 instance. Refinement and enhancement: By adding fresh data and adjusting the model's parameters, the model can be continuously enhanced and refined. The model's performance is improved and it can adjust to changing patterns in the text data thanks to this iterative procedure. Classification module: This module allows us to apply all conventional methods to feature data and determine the correctness of each method, as well as the algorithm's accuracy, precision, recall, and F Score. In the graph above, the numbers are plotted on the y-axis and the algorithm name is on the x-axis, displaying the accuracy, precision, recall, and f score for each method. It is crucial to remember that although sentiment and text analysis might offer insightful information, they shouldn't be relied upon as the only diagnosis method for VIRUS disorders. In order to identify diseases more precisely, these methods can be used in conjunction with other medical and diagnostic procedures including laboratory testing and clinical assessments.
The use of diverse methods and technology for VIRUS detection and analysis in the healthcare industry has significant benefits. The following are some main advantages:
Early identification: Health care providers can discover cases of COVID-19 even before patients show significant symptoms because to effective detection and analysis techniques. Early discovery slows the virus's transmission and may even save lives by allowing for prompt intervention, isolation, and suitable treatment. Making decisions with knowledge: The detection and study of VIRUS, driven by data, offers significant insights into the disease's epidemiological patterns, transmission rates, and severity. By using these findings, healthcare authorities can lessen the impact of the virus by making well-informed decisions about resource allocation, public health initiatives, and policy execution. Better treatment plans: Researchers and medical practitioners can create and improve treatment plans by analyzing COVID-19 data, which includes patient symptoms, comorbidities, and treatment outcomes. Healthcare professionals can customize interventions and enhance patient outcomes by knowing the elements that affect the course of an illness and how well a patient responds to treatment. Monitoring and surveillance: Tracking the virus's progress and evaluating its effects on public health depend heavily on VIRUS detection and analysis. Authorities can execute targeted interventions and preventive actions by identifying hotspots, developing variations, and prospective epidemics through real-time data analysis. Telemedicine and remote monitoring: These choices are becoming more and more feasible with the aid of VIRUS detection and analysis tools. Remote patient submission of health-related data, including symptoms and vital signs, eliminates the need for in-person trips to medical institutions. This method reduces the chance of exposure while also making it easier for medical professionals to monitor and evaluate patients' situations. Investigation and advancement: The examination of VIRUS data supports current investigations and advancements. Researchers can learn more about the behavior of the virus, how it affects distinct demographic groups, and how successful alternative treatments are by looking at large-scale statistics. This information helps in the creation of treatments, vaccinations, and public health campaigns to successfully battle the pandemic. Overall, early identification, well-informed decision-making, resource management, and enhanced patient care are all made possible by VIRUS detection and analysis in the healthcare industry. These strategies are essential for controlling the virus's influence on healthcare systems, preventing its transmission, and ultimately saving lives.
5 Claims & 1 Figure
Brief description of Drawing
In the figure which are illustrate exemplary embodiments of the invention.
Figure 1, The Process of identifying VIRUS using Classification-Based Algorithms. , Claims:The scope of the invention is defined by the following claims:

Claim:
1. A Classification Based system/method for VIRUS detection that consists of the following steps:
a) The Patient data is entered into the system at the beginning (1), and data collection is then carried out (2).
b) The patient's data is entered into the dataset construction (3). The undesirable attributes will then begin to be eliminated from the dataset by the data preparation step (4).
c) The different characteristics are extracted via feature engineering (5) depending on the need. The VIRUS variation (7) is predicted by using classification-based methods to the categorization (6).
2. As stated in claim 1, the patient information is gathered through the data collecting method in order to create new datasets.
3. As stated in claim 1, the superfluous data is eliminated from the dataset by data preparation.
4. In accordance with claim 1, the dataset's features are extracted using the feature engineering process, which also computes the probabilistic values and scores.
5. In accordance with claim 1, the classification-based method applies several categorization algorithms to forecast the VIRUS and enhances prediction accuracy.