Table of ContentsData Collection and Data Cleaning of Data SourcesData AnnotationMachine Learning Tools and Classification AlgorithmsResearchers Used a qualitative research method to collect and analyze necessary data on perception, expectations and concerns in LMS. The questionnaire used was formulated and validated by a social science researcher and is composed of two (2) open questions. These questions were asked directly to find out the feelings of students and teachers regarding the use of LMS as an educational tool. The data collected by the researchers totaled 1,321 responses on the user's perception, concerns and expectations regarding the educational uses of the LMS. The methodological structure of this study is composed of data collection, data cleaning, data annotation via manual classification of perception into positive or negative and machine learning of the training set shown in Figure 1. The researchers manually classified the collected responses as negative. perception or positive perception. To test the performance evaluation of the classification model, the researchers used 10 stratified cross-validations with a support vector algorithm applied in machine learning tools. One of the fastest growing technical fields today is machine learning, which sits at the intersection of computer science and statistics and at the heart of data science. In this study, it is used to validate the performance of the data model and then the expectations and concerns are categorized. thanks to thematic modeling [10]. Say no to plagiarism. Get a tailor-made essay on “Why Violent Video Games Should Not Be Banned”? Get the original essay Data collection and data source The survey questions were published online via Google Docs/Sheets. The researchers gave students and faculty the opportunity to answer two (2) open-ended questions: what are your perceptions of using the LMS as an educational delivery tool; and What are your expectations and concerns regarding using the learning management system as a learning tool. The respondents' data and their responses were recorded in a spreadsheet for the data cleaning process. Data CleaningSince the data collected is noisy, researchers undergo the process of data cleaning in the form of removing duplicates, symbols, numbers, and capital words. To remove duplicates, the machine learning tools that were used are Waikato Environment for Knowledge Analysis (WEKA). Notepad++ was used to process symbols, uppercase, lowercase via regular expressions. After data cleaning, only 770 perception responses were retained. Data Annotation Researchers manually categorized the collected perception response data into positive or negative perceptions, while expectations and concerns will be determined through thematic modeling using using the mallet tool. Negative answers classified by the researcher if the answer has negative auxiliary words such as "don't do", "don't", etc. Table 1 shows the example of positive and negative answers classified cleaned: Positive answer Negative answer useful tool many topics covered allow in-depth knowledge to be acquired no good educational tool allows transparent dissemination of information about the 2.