<p class="Abstract" style="text-align: justify;"><em>Abstract</em>— Speech recognition is a system to transform the spoken word into text. Human voice signals have a very high of variability. Speech signals in the different pronunciation text, also resulting in distinctive speech patterns. This, furthermore, happens if the text is spoken by a speaker who is not the mother tongue of the speakers. For example, text Arabic words spoken by Indonesian speaker. In this study, Mel Frequency cepstral Coeffisients (MFCC) feature extraction techniques explored for voice recognition of the Arabic words for Indonesian speakers with data training using Arabian native speakers. Furthermore, features that have been extracted, classified using Hidden Markov Model (HMM). HMM is one of the sound modeling where the voice signal is analyzed and searched the maximum probability value that can be recognized, from the modeling results will be obtained parameters are then used in the word recognition process. Recognized word is a word that has the maximum suitability. The system produces an accuracy by an average of 83.1% for test data sampling frequency of 8,000 Hz, 82.3% for test data sampling frequency of 22050 Hz, 82.2% for test data sampling frequency of 44100 Hz.</p>

This study aims to examine the effect of problem-based learning and tacit knowledge on problem-solving skills when students study in the laboratory. The method employed in this research was Quasi-Experimental Design. Data collection techniques were questionnaires and tests. Seventy-seven students were taken as the research participant and divide into two groups; 35 students in the experimental group with problem-based learning implementation and 42 students in the control group with procedural instruction. Questionnaires were used to measure tacit knowledge adopted from Insch, McIntyre, & Dawley (2008), Chilton & Bloodgood (2007), Somech & Bogler (1999). Data analysis techniques used two-way ANOVA test to determine learning outcomes. The research found that problem-based learning has a significant effect on problemsolving skills, and the use of tacit knowledge depends on the learning model. The results showed that problem-based learning could improve the ability of problemsolving while learning outcomes indicate that students use their tacit knowledge for problem-solving.

Piles of waste increase in line with population growth and consumption patterns. The concept of bioconversion using black soldier fly larvae can solve the problem of organic waste management. From these problems, an application of Internet of Things technology is needed. The system implemented aims to allow the system to find out how much accuracy, precision, and recall are in making decisions on media quality values using the Naive Bayes method. The main feature of this Naive Bayes Classifier is the very strong assumption of the independence of each condition or event. From the research results, the system has been successfully built according to the research design, as well as the goals that have been fulfilled in completing the development of the smart maggot. Several sensors used in this study were tested so that sensor performance could be determined by finding the average error value. Three parameters are measured; namely, the temperature obtained an average error of 1.6%, air humidity obtained an average error of 2.03%, and soil moisture obtained an average error of 2.7%. By measuring using Python, the Confusion Matrix is obtained so that the test results from the calculation of the Naive Bayes method can find the data in the form of accuracy, precision, and recall. Accuracy percentage results obtained 92%, precision percentage average results obtained 93%, and recall percentage average results obtained 92%. The conclusion shows the results of the system's accuracy obtained have worked well.

Government builds public facilities to support the needs of the community. The use of these public facilities needs to be re-evaluated, and one way to do it is through community response. Google Maps is one platform that receives the most responses from the community about location. Google Maps Reviews allow us to see how the public reacts to a location. Naïve Bayes method is used for classification in this study because it is one of the simple methods in machine learning that can be easily applied to several experiments conducted by the author. In the classification process, reviews produce many features that will be calculated based on their class. More features generated, more features processed too in the system. Chi-Square feature selection will be used to reduce features that have low dependence on the system. In this study, performance values will be calculated based on the experimental use of feature ratios of 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, and 100%. The results show that the use of 10% Chi-Square features produces the best performance, with an accuracy rate of 86.94%, precision of 80.42%, recall of 80.42%, and f-measure of 80.42%.

This study presents an analysis of several factors related to Indonesian student achievement. The sample consists of one hundred and eleven students in the second semester at the State Islamic University of Malang in Indonesia. The sample represents 74% of the student population for the even semester of 2021. Data were collected through instruments consisting of internal and external factors consisting of seventy-six statements. The age of the students ranged between 19 and 20 years. Internal factors which were analyzed included health, intelligence, talents, interests, motivations, and student learning methods. External factors analyzed include lecturers, other students, facilities, courses, extracurricular activities, and achievements. Then, internal and external factors were analyzed using multiple linear regression. These two factors simultaneously affect student achievement. Partially, the internal factors that influence learning achievement are the variables of intelligence and talent. Meanwhile, partially external factors that dominate the influence of learning achievement are lecturers and curriculum variables.

<p>Sistem pendukung pembelajaran tenses adalah suatu sistem pembelajaran terkomputerisasi yang dirancang untuk membantu seseorang dalam mendalami tenses. Tenses merupakan materi dasar dalam grammar yang digunakan untuk menunjukkan waktu kejadian memiliki struktur kata pembentuk kalimat yang berbeda-beda pada tiap penunjuk kejadiannya. Dari struktur kata pembentuk kalimat itulah sistem dapat membedakan tenses bentuk apa kalimat tersebut. Untuk membedakannya, sistem menggunakan metode forward chaining. Metode forward chaining adalah metode yang digunakan untuk mencari kesimpulan dari fakta-fakta yang terkumpul. Sistem kerja aplikasi ini adalah dengan memecah susunan kalimat menjadi kata kemudian dari kata tersebut oleh sistem akan dicari fakta-fakta dari kata tersebut. Fakta-fakta tersebut adalah jabatan berupa subjek, objek, verb dan lain sebagainya. Dari fakta-fakta tersebut, pada tahap akhir sistem akan mencari kecocokan antara fakta-fakta pembentuk kalimat dengan rumus pembentuk tenses. Dari hasil penelitian dengan memasukkan kalimat-kalimat yang memiliki struktur yang berbeda, aplikasi ini mampu mengenali bentuk tenses pada kalimat-kalimat tersebut. Hal ini mengacu pada hasil pengujian yang didapatkan persentase sebesar 96% dari 100 kalimat masukan.</p><p><strong>Kata Kunci</strong> :<em> </em>sistem pakar, forward chaining, tenses</p>

No abstract

Higher education institute should have maintained its student’s success in their academic field. Therefore, higher education institute should make a model to predict its success as early as possible. To have such model, the impacting factors should be determined, which factor are in the forms of continuous dan categorized data. This research is aimed to build logistic regression model based on compounded data from continuous and categorized data, which then test the model accuracy to predict the student’s success in such institute. Research data is using 68 student’s data. There are 6 research steps, first, data preparation and collection, second, data analysis, third, building the logistic multinomial regression model, fourth, data testing and validation, fifth, measure the model accuracy, and the sixth step is drawing the conclusion based on the analysis output. The result of prediction analysis dan accuracy test using logistic multinomial regression is the best model with significant factors that influence the study time are gender, department and selection track, while prediction accuracy model for each study time response variable is 96.4%.