DSpace Repository :: Browsing by Author "Mahdiyah, Evfi"

Browsing by Author "Mahdiyah, Evfi"

Now showing 1 - 20 of 30

ANALISA PERBANDINGAN PERFORMA ALGORITMA CONVOLUTIONAL NEURAL NETWORK DALAM KLASIFIKASI GAMBAR ABJAD BAHASA ISYARAT INDONESIA
(Elfitra, 2022-07) Kamandanu, Kamandanu; Mahdiyah, Evfi
Background clutter is a background in an image that has a lot of noise or objects, making it difficult to focus on one object. Generally, images with a cluttered background are more common than images with a green background. The purpose of this study is to see whether the background image affects the performance of the convolutional neural network algorithm in classifying the Indonesian Sign Language Alphabet (BISINDO). This deep learning model uses 2860 images of primary data for each background with a total dataset of 5720 images, the data that has been collected is divided into training, validation and testing, 3x3 filter size, and a learning rate of 0.001 and 50 epochs. of 0.983, validation of 0.823 and testing of 0.67 for model 1 (Green Background), while for model 2 (Background Clutter) the training accuracy is 0.971, validation is 0.529 and testing is 0.38. It can be concluded tha a pictures with background clutter affects the accuracy of the model.
ANALISA PERSEPSI MASYARAKAT TERHADAP DRAF KUHP 2019 MENGGUNAKAN METODE K-NEAREST NEIGHBOR
(2020-06) Ramadhan, Rizki; Mahdiyah, Evfi
The drafting of the new National Criminal Code (KUHP) for the discussion of the Criminal Law Act The legacy of the Dutch colonial government with all its amendments is one of the efforts in the development of national law. The government agreed to draft a new Penal Code, but reaped pros and cons in the community. Rejection of the design is not only done through social campaigns but also through social media Twitter so that it becomes a trending topic. The purpose of this study was to determine the number of perceptions of pros and cons on Twitter social media by using the K-Nearest Neighbor classification method. The data used are tweets containing #RKUHP and #RUUKUHP from September to October 2019. The results of this study were obtained 81% of people who had a Counter income and 19% of people who were pro, compared to the draft of the Criminal Code 2019. From the results of the study using matrix confusion, precision The highest system obtained was 77.16% which K = 1, and the highest recall value was 93.15% which K = 5 and K = 7.
ANALISIS DATA MINING UNTUK MENENTUKAN STRATEGI PENJUALAN PRODUK MENGGUNAKAN ALGORITMA FP-GROWTH (STUDI KASUS : MAZAYA OLEH-OLEH PEKANBARU)
(Elfitra, 2023-11) Muliyani, Sri; Mahdiyah, Evfi
Mazaya Souvenir Shop Pekanbaru is a company that operates in the culinary sector to be ranked first in the largest culinary industry in Pekanbaru, starting from processing, raw materials, packaging quality and distribution activities. This shop is experiencing difficulties in determining market strategies, one of which is in arranging products purchased simultaneously. The aim of this research is to carry out data mining analysis to determine product sales patterns using the Fp-Growth algorithm. The data used to carry out calculations using the Fp-Growth algorithm must go through several stages such as preprocessing the data, determining minimum support and confidence, determining the frequent itemset header, forming an fp-tree, creating a conditional pattern base based on the fp-tree, and generating frequent itemsets. In this study, a minimum support value of 1% and a confidence value of 80% were used. The research results showed that there were five rules, two frequent itemsets that had a confidence value of ≥80%, namely the products Risol Mayo Lumer and Talam Durian, Risol Ayam and Talam Durian, Red velvet and Talam Durian, Talam Durian and Risol Ayam, Talam Durian and Risol Mayo Lumer.
ANALISIS KEBERHASILAN MEDIA SOSIAL SEBAGAI MEDIA PROMOSI PADA PELAKU BISNIS ONLINE MENGGUNAKAN METODE TECHNOLOGY ACCEPTANCE MODEL (TAM)
(Elfitra, 2023-05) Wahyuni, Tri; Mahdiyah, Evfi
The purpose of this study was to determine the success rate of social media as a promotional medium for online business development. Research data were collected by means of literature studies and questionnaires. The questionnaire uses the online-based Google form service. The devices used are hardware and software. Data analysis was performed using multiple linear regression methods. The results of the analysis of the success rate of Instagram social media as a promotional medium for online business people using the Technology Acceptance Model (TAM) method can be concluded that by taking respondents twice, where the initial respondents were taken as many as 25 people in order to find out the validity test with valid results and test Reliability with Reliable results. Then the final respondent was taken again as many as 100 respondents in order to find out the average value for each variable and it is known that 63,9% of the value indicates that Behavioral Intention in Use (BITU) is influenced by Perceived Ease of Use (PEOU), Perceived Usefulness (PU), Attitude Towards Using (ATU), while the remaining 26.1% is influenced by other factors not examined in this study.
ANALISIS KEPUASAN PENGGUNA E-WALLET DANA MENGGUNAKAN METODE EUCS PERSPEKTIF PENGGUNA E-WALLET DANA KOTA PEKANBARU
(Elfitra, 2022-10) Rizky, Mutia Donda; Mahdiyah, Evfi
User satisfaction can be said to be one of the priority factors determining the success of an e-wallet application so that e-wallet provider companies must pay attention to supporting things on the system or application that can add to user satisfaction. This study was aimed at analyzing the satisfaction of DANA e-wallet users. The purpose of this study was to find out the extent of user satisfaction, the factors that user complaints, and recommendations to the DANA e-wallet to be able to increase user satisfaction. The EUCS (End User Computing Satisfaction) method can analyze factors that affect user satisfaction based on user experience in using the system by paying attention to the dimensions of content, accuracy, format, ease of use, and timelines. The data collection technique in this study was to use a questionnaire that was distributed to 349 respondents of e-wallet users in Pekanbaru City. The analysis technique carried out in this study was using the User Satisfaction Index (IKP) and the Importance Performance Analysis (IPA) Method. The results of this study, namely the overall level of user satisfaction which can be seen from the IKP value with a percentage of 81,18%. It shows users are very satisfied with the performance of the DANA e-wallet. In addition, none of the factors that are complaints for users can be seen from quadrant I of the results of the IPA Method. This research also provided recommendations to the DANA e-wallet in accordance with the results of the IPA Method analysis.
ANALISIS KUALITAS LAYANAN E-GOVERNMENT MENGGUNAKAN METODE E-GOVQUAL DAN IMPORTANCE PERFORMANCE ANALYSIS
(Elfitra, 2022-01) Julian, Ahmad; Mahdiyah, Evfi
The existence of information technology in government is expected to provide optimal service quality for the society. Based on Presidential Instruction Number 3 of 2003 which explains that every government is obliged to use the system and carry out evaluations continuously to achieve good management of IT or good e-Government. Based on the presidential instruction, the Department of Communication and Information Technology of Rokan Hilir Regency took the initiative to develop public services in the form of a website to improve work procedures and management systems within the government by utilizing information technology. This study aims to analyze the service quality of the website with the rohilkab.go.id domain based on the perception or point of view of its users using the e-Govqual and IPA methods, determine the factors that affect service quality based on the e-Govqual method and determine the attributes that are the priority for quality improvement. services based on the results of IPA analysis. From the 106 respondents who were used as research samples, the results of the e-Govqual analysis on the level of conformity have an average value of 83.18% and the value of the gap level has a GAP value of 0 which is -0.73, meaning that the performance of website services has not been able to meet user expectations. Furthermore, from the results of the quadrant analysis, obtained 9 attributes with the main priority, namely the attributes EU4 (configuring links with search engines), EU6 (personalization of information), FE1 (the presence of an online help form), RL3 (timely service), CA1 (complete data and information), CA3 (data relevance), CA4 (updated information), CA5 (all links are working properly), and CS5 (contact information), for service improvement
ANALISIS SENTIMEN MASYARAKAT TERHADAP KEPUASAN BELAJAR DALAM JARINGAN (DARING) DI MASA NEW NORMAL MENGGUNAKAN METODE NAÏVE BAYES CLASSIFIER
(perpustakaan UR, 2021-10) Kumara, Roby; Mahdiyah, Evfi
Online learning is a method of studying over the internet. Noawadays in Indonesia began to implement the policy of online learning by utilizing learning management system application as a learni006Eg medium,as one of the efforts to prevent the widespread transmission of corona virus(Covid-19) in the world of education. This policy of online learning activities remotely began to be widely discussed not only through the real world but also on cyberspace, through social media such as twitter so that it became a trending topic to be discussed. The purpose of thisstudy is to find out the percentage of Pros and Contras on twitter’s users by using the one of the classification method, Naïve Bayes Classifier. The data used for this study are tweets that contains #belajardaring and #belajardarirumah from September to October 2020 period. When collecting data using twitterscraper that is loaded using pyhton. Then pre-processing the data that includes case folding, tokenizing, normalization, stopword removal, and stemming. The labeling process resulting two classes of data, positive and negative, with a total of 417 data. The results of this study foundthat 69% of people's perception is Contras and the other 31% is Pros, referring the online learning 2020 issues. Data that already has a class is then divided (Split) by 70% training data and 30% testing data to be used in the Naïve Bayes Classifier algorithm. The evaluation showed an accuracy score of 79.63%, a recallvalue of 88.49% and a precision value of 66.66% based on 417 tweets data,consisting 291 training data and 126 test data
ANALISIS SENTIMEN TERHADAP PELAKSANAAN PEMBATASAN SOSIAL BERSKALA BESAR (PSBB) PADA MASA NEW NORMAL MENGGUNAKAN METODE SUPPORT VECTOR MACHINE (SVM)
(Elfitra, 2022-01) Handayani, Dwi; Mahdiyah, Evfi
The spread of the COVID-19 virus has now become a pandemic because it has spread throughout the world, including Indonesia. The prositive case rate and the increasing death rate led the Indonesian government to make a regulation limiting community activities to break the chain of spread of COVID-19 in Indonesia. This regulation is a Large-Scale Social Restriction (PSBB. This becomes a polemic for some Indonesian people, especially those who do not have a fixed income. Many social media users discuss the implementation of PSBB in Indonesia. This study aims to find out how public’s sentiment towards the implementation of PSBB in Indonesia during the new normal period through tweets on social media Twitter using the Support Vector Machine (SVM) method. The data obtained for the implementation of the PSBB before the new normal was 322 tweet data that had been labeled while the tweet data for after the new normal was 653 data that had been labeled. The results obtained using the k-fold cross validation and confusion matrix method on the model created results in the highest accuracy rate is 81% with the distribution of data 90% : 10% for data before the new normal and for the data after the new normal, the highest accuracy value is 71% with the distribution of data 80% : 20%. The use of k-fold cross validation produces the highest accuracy value of 78% with fold value = 8 for data before new normal, while for data after new normal, the highest accuracy values is 73% with fold value = 1.
ANALISIS SENTIMEN TWITTER TERHADAP PEPERANGAN RUSIA DAN UKRAINA MENGGUNAKAN ALGORITMA SUPPORT VECTOR MACHINE
(Elfitra, 2022-07) Andraini, Fitri; Mahdiyah, Evfi
Sentiment analysis is one of the sciences in Text Mining which is used to classify text documents in the form of public opinion. The text document used in this study comes from Twitter using three positive, negative, and neutral labels. The method used was the Support Vector Machine. This study used 2000 data using two data test scenarios, namely, Split Validation and Cross Validation. Testing data with Split Validation was divided into 4 variations, namely, 50%:50%, 70%:30%, 80%:20%, and 90%:10%, while testing data with Cross Validation was divided into 10 tests. The highest accuracy result used the Split Validation test laid in a comparison of 90%: 10%, which is 83.5%, while data testing with Cross Validation had the highest accuracy at K-Fold 7 of 86.0%.
APLIKASI PREDIKSI STOK ALAT KONTRASEPSI STUDI KASUS BKKBN PROVINSI RIAU MENGGUNAKAN METODE LEAST SQUARE
(Elfitra, 2023-09) Wahyuda, Alvin; Mahdiyah, Evfi
The importance of data collection in determining the amount of stock of contraceptives, makes BKKBN need to predict and record the outgoing stock and incoming stock of contraceptives so that forecasting is needed in predicting the number of contraceptives. The aim of this research is to predict the number of stocks of contraceptives using the least squares method. The Least Square method is a time series data method that requires historical data to anticipate future sales in order to determine the results. The system design uses the Unified Modeling Language that are Use Dase Diagrams, Sequence Diagrams and Class Diagrams which are implemented using the PHP programming language and MySQL database. The data used in the study are the expenditure data for contraception devices at BKKBN Riau Province from January to December 2022. The predicted results of contraceptive stock (least square/Y^) in January 2023 are the types of birth control pills (10,347), implants (3,375), progestin pills/Gestin F3 (247), condoms (80), progestin injections (23,555) and IUDs. Chopper T (543). Prediction results were tested using MSE for the types of birth control pills (8,921,648), implants (949,304), progestin/gestin F3 pills (5,083), condoms (533), progestin injections (46,238,287) and Choper T IUD (24,587). The prediction results were declared successful because the least square (Y^) value obtained was lower than the MSE value, so this system can be recommended to related parties who need it.
EVALUASI USABILITY WEBSITE MENGGUNAKAN METODE SYSTEM USABILITY SCALE (STUDI KASUS SEKOLAH SMAN 9 KECAMATAN MANDAU)
(Elfitra, 2022-01) Adnan, Zaki; Mahdiyah, Evfi
SMAN 9 Mandau has provided website to improved quality of information dissemination. Evaluation of the SMAN 9 Mandau website needs to be done, especially from the aspect of user convenience. The purpose of this study was to measured the level of user satisfaction on SMAN 9 Mandau Website using System Usability Scale (SUS) method. This method has 10 items and has 5 Likert scales. SMAN 9 Mandau consists of 858 people. This study used slovin technique to determined number of respondents' samples and results obtained are 272 respondents. Data was collected by distributing questionnaires to respondents using Googleform. Based on the results, the SUS score was 63, Adjective Rating was in the OK category with a grade scale of C- and the acceptability level was Marginal. It means that website has been accepted, but the acceptance is relatively low, so it is necessary to improvements of the website.
EVALUASI USABILITY WEBSITE MENGGUNAKAN METODE SYSTEM USABILITY SCALE (STUDI KASUS WEBSITE UNIVERSITAS RIAU)
(Elfitra, 2022-07) Surianda, Panggi; Mahdiyah, Evfi
The application of information technology at this time is increasingly important because it can provide various benefits for the University. The website involves many users such as students, lecturers, guardians and others, but because currently technology is very advanced, a website should be user friendly so that users can easily use the features provided by the website. This study uses the system usability scale (SUS) method to improve the Riau University website, therefore. System usability scale is a technique used to measure how effective a website is for users by collecting opinions from people involved in operating the website. This study aims to measure the level of user satisfaction on the Riau University website using the System Usability Scale method and determine the Acceptability, Grade Scale and Adjective Rating values on the Riau University website. The research process starts from determining the number of samples, then distributing the SUS questionnaire, data processing is carried out, then calculating the SUS score with a result of 74,89 which is included in grade B, the last is a system design using Unified Modeling Language (UML).
IMPLEMENTASI METODE TREND MOMENT DALAM MERAMALKAN HARGA MINYAK GORENG DI 9K SWALAYAN
(Elfitra, 2023-07) Fildzadeno, Sy. Salsabila; Mahdiyah, Evfi
Setting the selling price is one of the factors that can be both profitable and detrimental for the seller. If you choose the wrong selling price, the trader can suffered big losses. Cooking oil, which is a staple in Indonesia, often experiences ups and downs, demanding that sellers be able to provide favorable prices for companies and buyers. The purpose of making this forecasting system is to find out whether the trend moment method can predict the selling price of Avena, Bimoli Classic and Fortune cooking oil at 9K Supermarkets properly, and implement it into a web-based system. Trend moment is the best calculation among other types of forecasting calculations because there is no difference between odd and even data. The system design uses the Unified Modeling Language (UML), there are use case diagrams, sequence diagrams and class diagrams. Implementation in the PHP programming language and MySQL Database. Forecasting is carried out using cooking oil selling price data for the period January 2021 to February 2022. The results of the cooking oil selling price forecasting that have been obtained in this study are Avena IDR 14,984.71, Bimoli Classic IDR 15,587.97 and Fortune IDR 15,700.32. The Mean Absolute Error (MAPE) for Avena is 0.50%, Bimoli Classic is 0.80% and Fortune is 0.86%, which are included in the very good forecasting criteria.
IMPLEMENTASI RATIONAL UNIFIED PROCESS (RUP) DALAM PERANCANGAN SISTEM INFORMASI AREA PARKIR KENDARAAN BERMOTOR PADA FMIPA UNIVERSITAS RIAU
(perpustakaan UR, 2021-07) Arianda, Muhammad Vieri; Mahdiyah, Evfi
The management of the FMIPA Riau University parking lot is still done manually, where every motorized vehicle that will park is only given a parking ticket sheet and no routine recording is carried out. The purpose of this study is to implement the Rational Unified Process (RUP) in the design of a motorized vehicle parking area information system at the Faculty of Mathematics and Natural Sciences, Riau University. This information system is built using the PHP and MySQL programming languages. The system development method used is the Rational Unified Process (RUP) method which is divided into three phases, namely inception, elaboration and construction. Each stage has several processes carried out starting from business modeling, requirements, design, implementation, and testing. The process carried out in the inception phase is the business modeling and requirements phase, then in the elaboration phase the design phase is carried out and the construction phase is the implementation phase. The final result of the design of this information system, can be used to make it easier to manage student motorized vehicle parking data at FMIPA UR and provide convenience to the safety of student motorized vehicles.
KLASIFIKASI ARTEFAK PADA GAME GENSHIN IMPACT MENGGUNAKAN ALGORITMA NAIVE BAYES CLASSIFIER
(Elfitra, 2023-01) Rochyadi, Fachri Romi; Mahdiyah, Evfi
One type of entertainment that people often do is playing games. Genshin Impact is a very popular game. In this game, there are many characters that can be obtained during the game. Players need to strengthen their character to defeat enemies during adventures and complete missions. One way to strengthen the character is to use artifacts. There are so many artifacts in Genshin Impact that each artifact has a random value that is confusing for new players. This study aims to classify artifacts using the Naive Bayes Classifier method. This Naive Bayes Classifier algorithm model uses a dataset of Gladiator and Wanderer artifacts of 349 data, 279 data as training data and 70 data as testing data. The data attributes used are 8 attributes, the name of the artifact set, the type of artifact, the type of main stat, sub stat 1, sub stat 2, sub stat 3, and sub stat 4. This study resulted in a classification model with an accuracy of 87.1% measured using the Confusion Matrix
KLASIFIKASI TWEET PADA TWITTER MENGGUNAKAN METODE MULTINOMIAL NAÏVE BAYES STUDI KASUS UNTUK TOPIK PRESIDEN
(Elfitra, 2022-01) Sandra, Nofri; Mahdiyah, Evfi
Social media invites anyone who is interested to participate by contributing and giving feedback openly, giving comments, and sharing information in a fast and unlimited time. One of the most widely used and influential social media in Indonesia is Twitter. Twitter usually contains meaningless chatter, conversation, repeated messages or retweets, selfpromotion, spam, news and some statements, opinions and expressions. Tweets can contain blasphemous or hate speech and Good Speech. This study aims to build an application to classify user tweets against posts on Twitter social media so that it is known whether the Tweet contains hate speech or not by implementing the multinomial nave Bayes method. Crawling data using the R programming language using the Rstudio tool. Use the Twitter API to search data with the keyword "president". The data is taken from comments containing keywords using a time filter, namely October 2018 - April 2019. The application is able to classify tweet data using multinomial nave Bayes with an accuracy of 87,5% based on 400 tweet data by dividing 360 training data and 40 testing data.
LOGIKA FUZZY TAHANI SISTEM PENDUKUNG KEPUTUSAN PEMILIHAN IKAN LAUT PADA PT KARUNIA SAMUDERA JAYA
(Elfitra, 2023-05) Siburian, Agustin; Mahdiyah, Evfi
Fish is a food that is produced from the sea and contains many good benefits for the human body. Often companies are overwhelmed in assessing the quality of fish, so that subsequent performance becomes sluggish. Freshness of fish is the main factor that determines the quality and durability of the product to be processed. Fish quality can be assessed from several organoleptic factors. In this digital era, a lot of research has been conducted to build a decision support system in assessing a case quickly and efficiently. This study aims to build a decision support system with fuzzy taught logic that is able to help find the best quality fish. The system that was built and developed uses Fuzzy Fuzzy Logic for the process of determining the best fish quality. This Fuzzy Logic has 3 stages, namely the membership function, fuzzification and calculating the highest score. The categories used are fish eyes, fish gills, surface mucus, fish odor, and fish texture. The fuzzy rules used in this study are to calculate the highest value of the quality of fish freshness. This research involved 50 fish samples at PT Karunia Samudera Jaya. The results of this study were fish that had the best quality in the categories of fresh fish eyes, fresh fish gills, fresh surface mucus, fresh smell and fresh texture found in box number 27 with a freshness level value of 0.92. Fuzzy Meghan Logic can be implemented into a decision support system for assessing the best quality of fish based on the condition of the criteria to be assessed on fish.
PENENTUAN LOKASI PARKIR DENGAN MENGGUNAKAN METODE FLOYD-WARSHALL PADA PARKIRAN BANDAR UDARA SSQ II BERBASIS WEB
(Elfitra, 2023-07) Paris, Nanda Toho; Mahdiyah, Evfi
The development of technology and information takes place very quickly, including in the field of Geographic Information Systems (GIS). The fast and rapid development of technology is very helpful for information to arrive quickly, precisely and accurately. GIS can provide information related to the location of an object or object that will be used to assist in making decisions, one of which is in determining the location of vehicle parking. The purpose of this research is to design and build an application that can determine the location of parking based on the closest distance so that it makes it easier for drivers to find parking locations. This research method uses the Floyd Warshall algorithm which searches for the closest distance by calculating all possible distances that can occur and will get the closest distance between two points by calculating all distances and will produce a route. The use of the Floyd Warshall algorithm can be used to make it easier for drivers to find the desired parking location and explain the route that must be passed in order to arrive at the destination. Based on the results of observations obtained data Sultan Syarif Qasim II Airport (SSQ II) has 10 blocks for four-wheeled vehicles. The result of this study is the closest distance from the starting point which is the entrance to the end point which is the closest destination location is in block B5.
PENENTUAN LOKASI SUPLIER GAS ELPIJI BERDASARKAN DATA LOKASI PENJUALAN ECERAN GAS ELPIJI MENGGUNAKAN METODE K-MEANS (STUDI KASUS KECAMATAN TAMPAN)
(Elfitra, 2022-05) Beti, Beti; Mahdiyah, Evfi
The density of population in an area is directly proportional to the amount of fuel used for household purposes, including LPG. Distribution of LPG gas must be carried out evenly to all regions, so that it is easier for people to get LPG gas. However, this is still not fully realized. The distribution of LPG gas by suppliers is still unable to meet consumer demand in several areas including the Tampan sub-district. This study aimed to provide recommendations for the location of construction of LPG Gas suppliers based on retail location data using the K - Means Centroid method. The process of research was started from collecting data on the coordinates of the LPG Gas sales location, after the data on the LPG Gas seller coordinates was obtained, a data preprocessing process was carried out, then the clustering process was carried out using the K Means Centroid Method The clustering process used a K value that varies from K : 3 to with K : 7 The best structure of this study was K: 3 with a test value of 0.67 which was classified in the medium structure cluster.
PREDIKSI LAPORAN PERKEMBANGAN PINJAMAN TABUNGAN PADA BADAN USAHA MILIK KAMPUNG MENGGUNAKAN ALGORITMA C4.5
(2020-01) Franciska, Annaa Adhe; Mahdiyah, Evfi
Prediction reports on the development of savings and loans are very necessary for the Village Owned Enterprise (BUM) of Anugrah Village. By using this prediction of the report on the development of savings loans, BUM Kampung Anugrah can predict the number of customers who repay loans on time and not on time. The purpose of this study is to predict reports on the development of savings and loans at BUM Kampung Anugrah. This prediction used the application of the C4.5 Algorithm and Decission Tree methods. This system was built using the PHP programming language with MySQL as the database server and UML as the system design. The results obtained from this study were in the form of prediction of savings loans that are on time and not on time as well. Finally, we performed the system testing using the Black Box method. This system helped the BUM Kampung Anugrah organization to manage efficiently customer loan data.

Browsing by Author "Mahdiyah, Evfi"

Results Per Page

Sort Options