OJS UNPATTI Publication Center (Universitas Pattimura)
Not a member yet
8549 research outputs found
Sort by
GOJEK DATA ANALYSIS THROUGH TEXT MINING USING SUPPORT VECTOR MACHINE (SVM) AND K-NEAREST NEIGHBOR (KNN)
Get PDFThe main focus of this research is to apply and test the effectiveness of SVM and KNN methods in Gojek data text analysis. This research will examine how the two methods can classify user comments and feedback and identify data sentiment analysis at the same time practically help Gojek understand user needs and improve service quality. The data obtained through scrapping is categorized into positive and negative sentiment. Data is taken from Gojek application user reviews throughout the year 2022 with a total of 1148 sentiment data with a percentage of 80% training data and 20% testing data. Evaluation of model performance using Confusion Matrix and AUC-ROC Curve shows that SVM is more effective than KNN, with accuracy on training data of 92.55% for SVM and 81.71% for KNN, as well as accuracy on testing data of 82.40% for SVM and 77,09% for KNN
ESTIMATING EARTHQUAKE MAGNITUDE USING SPATIAL INTERPOLATION WITH THE INVERSE DISTANCE WEIGHTING AND ORDINARY KRIGING APPROACH
Get PDFThe West Java region is known for its high disaster vulnerability, particularly to earthquakes, due to the presence of many active faults. Based on this, information is needed as an initial step for disaster mitigation to reduce disaster risk and guarantee community safety in the West Java region, one of which is by carrying out spatial interpolation. In this study, the Inverse Distance Weighted (IDW) and Ordinary Kriging (OK) methods were used. The research sample included data on earthquake magnitude in West Java within the same coordinate range, from 01 January 2022 to 31 December 2022. This research was conducted to find out a more precise spatial interpolation method in estimating the strength of the earthquake in West Java in 2022. From the OK analysis results, the best theoretical semivariogram model was obtained, namely the Exponential model with nugget, sill and range values of 0.07, 0.12 and 11451 meters. From the results of the IDW analysis, the best power value parameter was obtained, namely 2. This research was conducted to develop a more precise spatial interpolation method for estimating earthquake strength in West Java in 2022. The OK method results indicated that most of the West Java region has the potential for earthquakes with a magnitude of around 1.5 to 4.0, while the IDW method suggested a potential magnitude of around 2.0 to 4.0. The potential for a high-magnitude earthquake is in Kp. Cileuley, Garumukti, Pamulihan District, Garut Regency, West Java. Based on the results of Holdout Cross Validation, the IDW method is the best for estimating earthquake magnitude in West Java, with a MAPE value of 17.8%. The IDW method's estimation of earthquake magnitude is superior to OK, with smaller MAPE and MAE values
THE INFLUENCE OF MODERATING FACTORS IN STUNTING: LOGISTIC PATH ANALYSIS OF ORDINAL DATA
Get PDFLogistic path analysis is used to analyze direct and indirect causal relationships between exogenous-endogenous variables with categorical data types. This study aims to apply logistic path analysis to ordinal categorical data and model the relationship between exogenous variables that affect nutritional status and physical status (stunting) in toddlers in Sumberputih Village, Wajak District. The data used is secondary data obtained from the results of filling out questionnaires in Sumberputih Village at the time of data collection in 2022-2023. The sample used in the study was 100 housewives who had toddlers. The sampling technique used was judgment sampling. However, the study only selected the variables of Birth Weight, Dietary Habits, Nutritional Status, and Physical Status (Stunting). The result of this study is that the variable of Birth Weight has a significant direct effect on Nutritional Status. The variable of Birth Weight has an indirect effect, and the total effect on Physical Status (Stunting) mediated by Nutritional Status is not significant. Meanwhile, the Diet variable has a significant direct effect on Physical Status. In addition, the Socioeconomic Condition variable can moderate the relationship between the Birth Weight variable and Physical Status. The diversity of data that can be explained by the model is 80.36%, while the rest is explained by other variables outside the model by 19.64%
JOINT DISTRIBUTION AND PROBABILITY DENSITY OF CLIMATE FACTORS IN KALIMANTAN USING NESTED COPULAS
Get PDFIn this study, we investigate the joint distribution of local and global climate factors in Kalimantan, Indonesia, using fully and partially nested copula models. The analysis focuses on capturing the dependencies between local factors (precipitation and the number of dry days) and global indices (ENSO and IOD). The methodology involves estimating the marginal distributions of each variable using goodness-of-fit tests, and then modeling the dependence structure between variables with a range of copulas. We used both one-parameter copulas, including Gaussian, Clayton, Gumbel, Joe, and Frank, as well as two-parameter copulas, such as BB1, BB7, and BB8, with rotations of 90°, 180°, and 270° applied to account for negative dependencies. Nested copula structures were employed to model multivariate dependencies, with fully nested and partially nested approaches used to capture interactions between all four variables. The results show that global climate indices, particularly ENSO and IOD, have a more substantial influence during the dry season, impacting drought conditions in Kalimantan. The copula method offers a flexible and efficient way to construct multivariate joint distributions, better representing complex climate relationships than traditional models. Future work could extend this approach to include additional climate variables and use real-time data for forest fire risk prediction
ON THE SECURITY OF GENERALIZED MULTILINEAR MAPS BASED ON WEIL PAIRING
Get PDFIn 2017, Tran et al. proposed a multilinear map based on Weil pairings to realize the Boneh-Silverberg scheme. They proposed an algorithm to evaluate the Boneh-Silverberg multilinear map and showed that it could be used to establish a shared key in multipartite key exchange for five users. They claimed their scheme was secure and computable in establishing a shared key between 5 users. Unfortunately, they did not prove that their scheme meets three additional computational assumptions proposed by Boneh and Silverberg. In this paper, with some computational modifications, we show that the algorithm proposed by Tran et al. does not satisfy three security assumptions proposed by Boneh and Silverberg. Therefore, every user involved in this multipartite key exchange can obtain the shared key and other users' secret values. We also show that the computation to obtain a shared key is inefficient because it requires a lot of computation and time
Analisis Peluang Kejadian Deret Hari Kering Selama Musim Tanam Efektif pada Periode El Nino di Pulau Ambon
Get PDFSpell analysis is one of the indicators that can be used to measure the level of vulnerability of a region to drought. This study aims to determine the chances of a series of dry days during the effective growing season during El Niño on Ambon Island. Data analysis was carried out with the following stages: (1) calculating rainfall probability at 75% using the ranking order method and potential evapotranspiration using the Penman-Monteith method in the Cropwat 8.0 Program package, (2) calculating land water balance and determining the effective growing season based on optimum soil water content, and (3) determining the years of El Niño occurrence and the chances of a series of dry days. The results of the study showed that the effective growing season on Ambon Island lasted for 7 months, from April to October. The El Niño event on Ambon Island during the period 1979–2023 tends to occur once every three years. The chances of a series of dry days ≥ 5 days during the growing season ranged from 56 to 88%; ≥ 10 days: 27 to 88%; ≥ 15 days: 7 to 80%; and ≥ 20 days: 0 to 47%.
Keywords: dry spell, growing season, El Niño, Ambon Islan
DINAMIKA BRAND SWITCHING DALAM ISU GLOBAL: PRODUCT KNOWLEDGE, RELIGIOSITY DAN ELECTRONIC WORD OF MOUTH
Get PDFUnilever menjadi salah satu perusahaan yang terkena dampak dari aksi boikot. Akibat dari fenomena Masyarakat Indonesia mulai membatasi diri dengan tidak lagi membeli produk – produk yang dianggap mendukung genosida dan memilih untuk berganti ke produk lain. Penelitian ini bertujuan untuk mengetahui pengaruh Product Knowledge (X1), Religiosity (X2) dan Electronic Word of Mouth (X3) terhadap Brand Switching (Y) pada produk Unilever. Populasi yang dipakai dalam penelitian ini adalah adalah konsumen atau orang – orang yang pernah memakai produk – produk dari Unilever, dengan total sampel yang diteliti sebanyak 100 sampel. Penelitian ini berlokasi di Kota Ambon. Teknik sampling yang digunakan dalam penelitian kuantitatif ini adalah teknik purposive sampling. Pengujian dilakukan dengan menggunakan IBM SPSS Statistic 23 dengan teknik analisis data menggunakan regresi linear berganda. Hasil penelitian menunjukkan bahwa Product Knowledge berpengaruh negatif dan tidak signifikan terhadap Brand Switching, sedangkan Religiosity dan Electronic Word of Mouth memiliki pengaruh positif signifikan. Berdasarkan hasil uji nilai R Square (R2) Product Knowledge, Religiosity dan Electronic Word of Mouth mempengaruhi Brand Switching sebesar 64,8% dan sisanya 35,2% dipengaruhi oleh variabel lain
Purchase Decision of Club Brand Mineral Water: The Influence of Price, Quality, and Promotion in Jabodetabek
Get PDFPurchase decision is a crucial aspect of marketing strategy, especially in the highly competitive mineral water industry. This study aims to analyze the influence of price, product quality, and promotion on the purchase decision of Club brand mineral water in the Jabodetabek area. The research employs a quantitative approach by collecting primary data through a questionnaire distributed via Google Forms. The study sample consists of 125 respondents who are Club mineral water consumers in Jabodetabek, selected using purposive sampling. Data analysis was conducted through respondent characteristic tests, descriptive statistical analysis, multiple linear regression analysis, normality tests, t-tests, F-tests, and hypothesis testing. The results indicate that price does not have a significant influence on purchase decisions, as evidenced by a significance value of 0,065, which exceeds the 0,05 threshold, and a t-value of 1,864, which is smaller than the t-table value (1,979). Conversely, product quality is proven to have a significant influence on purchase decisions, with a significance value of 0,006 and a t-value of 2,822, which exceeds the t-table value. The promotion factor also has a significant influence, with a significance value of 0,000 and a t-value of 5,036. These findings suggest that consumers prioritize product quality and the effectiveness of promotions over price when making purchase decisions for Club brand mineral water. The novelty of this study lies in its specific analysis of consumer preferences within the mineral water product category in Jabodetabek, as well as the role of promotion as a more dominant determinant of purchase decisions compared to price. The implications of this study provide recommendations for marketing strategy developers, particularly in the bottled beverage industry, to focus more on quality improvement and promotional innovation to attract consumers. From a social sciences and humanities perspective, these findings contribute to consumer behavior studies within the context of microeconomics and value perception-based marketing strategies
Navigating Maritime Digitalization: Socioeconomic and Organizational Perspectives on Technological Transformation
Get PDFMaritime digitalization has become a transformative phenomenon that not only reshapes the operational landscape of the maritime industry but also has profound socioeconomic and organizational implications. This study aims to explore the impact of technological disruption on maritime management by examining digitalization dynamics through the lenses of disruptive innovation theory and the resource-based view. Employing a mixed-methods approach, this research integrates quantitative data analysis from industry reports with qualitative insights obtained through interviews and case studies of leading maritime enterprises. The findings reveal that digital technologies such as the Internet of Things (IoT), blockchain, and artificial intelligence significantly enhance operational efficiency and competitive positioning. However, the adoption of these technologies also presents complex challenges, including regulatory barriers, cybersecurity threats, and the need for workforce reskilling. This study proposes a conceptual model that illustrates the relationship between digital technologies, organizational capabilities, and market performance, providing a foundation for a deeper understanding of maritime digitalization dynamics. The novelty of this study lies in its approach, which not only highlights the technical aspects of digitalization but also examines how technological transformation affects the socioeconomic structure and work patterns within the maritime industry. The findings offer strategic recommendations for policymakers and maritime managers to navigate digital change effectively. Furthermore, this study contributes to the development of social sciences and humanities by emphasizing the interconnection between technological innovation, social change, and organizational adaptation in a historically conservative industry. Aligning technological investments with organizational strategies is a key factor in the successful digitalization of maritime operations, paving the way for future research on optimizing digital integration in maritime business management
MENINGKATKAN HASIL BELAJAR MATEMATIKA MAHASISWA MELALUI PEMBELAJARAN GUIDED DISCOVERY LEARNING DENGAN MEDIA PEMBELAJARAN CLASSPOINT
Get PDFThis study is a classroom action research conducted in two cycles, aimed at improving mathematics learning outcomes among first-semester students of the Agribusiness Study Program at the University of West Sulawesi. The research implemented the Guided Discovery Learning (GDL) model using ClassPoint as a learning medium. The success of the study is reflected in the achievement of the predetermined indicators. There was a noticeable improvement in learning outcomes: the average score in Cycle I was 67 with a learning mastery rate of 38%, while in Cycle II, the average score increased to 76 with a classical mastery rate of 83%. This indicates an increase of 9 points in the average score and a 45% improvement in learning mastery. Student learning responses also showed progress, rising from 69% in Cycle I to 85% in Cycle II—an increase of 16%. Additionally, student engagement in learning improved, with active participation rising from 70.6% in Cycle I to 76% in Cycle II. Lecturer activity and classroom management were also positively evaluated, with observation scores increasing from an average of 71.3% in Cycle I to 82% in Cycle II