Garuda - Garba Rujukan Digital

Implementation of Global Vectors for Word Representation (GloVe) Model and Social Network Analysis through Wonderland Indonesia Content Reviews Singgalen, Yerik Afrianto
Jurnal Sistem Komputer dan Informatika (JSON) Vol 5, No 3 (2024): Maret 2024
Publisher : STMIK Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v5i3.7569

Integrating the Global Vectors for Word Representation (GloVe) Model with Social Network Analysis presents a promising approach for extracting nuanced semantic relationships from Wonderland Indonesia's content reviews. However, the lack of comprehensive studies exploring the effectiveness of this integration, specifically within the context of Wonderland Indonesia's content reviews, necessitates focused research to uncover its potential impact and applications. This study investigates the reception and impact of the "Wonderland Indonesia" video content by Alffy Rev ft. Novia Bachmid (Chapter 1) within the YouTube community using a comprehensive methodology based on CRoss-Industry Standard Process for Data Mining (CRISP-DM), topic analysis, and Social Network Analysis (SNA). Through topic analysis, the content's main themes and narrative elements were identified, shedding light on its storytelling effectiveness. Furthermore, sentiment analysis using Vader was conducted on 2204 out of 24185 posts, revealing that 1369 (92%) exhibited positive sentiment, 427 (31.19%) had neutral sentiment, and 850 (62.09%) contained negative sentiment. Additionally, sentiment analysis using TextBlob was performed on the same subset of posts, with 1369 (40) posts exhibiting positive sentiment, 599 (43.75%) with neutral sentiment, and 730 (53.32%) expressing negative sentiment. Notably, metrics such as toxicity (highest value: 0.90780) and severe toxicity (highest value: 0.95021) exhibited varying prominence within the analyzed content. These findings enable targeted interventions and content moderation strategies to promote healthier online discourse. The SNA uncovered intricate social dynamics and interaction patterns among viewers, emphasizing the video's ability to foster engagement and community interaction. This study underscores the significance of creative storytelling and community engagement strategies in digital content creation, with implications for audience participation and community development within the digital sphere. Future research could explore the longitudinal effects of such content strategies on audience retention and community engagement.

Sentiment Classification of Food Influencer Content Reviews using Support Vector Machine Model through CRISP-DM Framework Singgalen, Yerik Afrianto
Jurnal Sistem Komputer dan Informatika (JSON) Vol 5, No 3 (2024): Maret 2024
Publisher : STMIK Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v5i3.7509

The research problem revolves around the challenges in effectively marketing culinary tourism aligned with tourist preferences in Indonesia, necessitating a substantial exploration of consumer sentiments related to culinary diversity through the lens of food influencer content. Food influencers are crucial in stimulating tourists' interest in gastronomy through culinary tourism in Indonesia. This research reveals challenges in culinary tourism marketing aligned with tourist preferences, necessitating substantial exploration of consumer sentiments related to culinary diversity through food influencer content. The sentiment classification method employed is the Cross-Industry Standard Process for Data Mining (CRISP-DM) using the Support Vector Machine (SVM) algorithm and the SMOTE operator. The data source is derived from a video with the ID PMhfLy_buV8, containing 114,422 comments. This study collects and processes 30,000 comments, resulting in 9,323 data points. The findings highlight the vital performance metrics of SVM models, both with and without SMOTE, showcasing high accuracy, precision, recall, and F-measure values. Specifically, SVM without SMOTE achieves 95.28% accuracy, while SVM with SMOTE achieves 98.67%. Despite some limitations in discerning positive and negative sentiments, indicated by moderate Area Under the Curve (AUC) values (0.608 to 0.658), the overall efficacy of SVM in sentiment analysis for food influencer content is apparent. Drawing from a dataset of 30,000 comments, these insights contribute to advancing sentiment analysis methodologies and offer practical implications for understanding consumer perceptions and behaviors in digital media and influencer marketing. Additionally, the prominence of frequent words such as "bang" (1322), "nonton" (1064), "makan" (921), "yang" (801), "puasa" (711), "tahun" (484), "ngiler" (448), "lagi" (384), "tanboy" (311), and "enak" (315), as extracted from RapidMiner analysis, underscores the significance of language patterns in the realm of food influencer content.

Toxicity Analysis and Sentiment Classification of Wonderland Indonesia by Alffy Rev using Support Vector Machine Singgalen, Yerik Afrianto
Jurnal Sistem Komputer dan Informatika (JSON) Vol 5, No 3 (2024): Maret 2024
Publisher : STMIK Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v5i3.7563

The music industry's increasing reliance on digital platforms like YouTube for dissemination raises concerns about the potential impact of music videos on viewer sentiment and well-being. This study seeks to assess the toxicity and sentiment of the Wonderland Indonesia music video by Alffy Rev through Support Vector Machine analysis, contributing to our understanding of the effects of music content on online audiences. This research addresses the challenge of sentiment classification in digital content by leveraging the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework. The study aims to enhance sentiment classification accuracy by applying a Support Vector Machine (SVM) with a Synthetic Minority Over-sampling Technique (SMOTE) to address class imbalance issues. The research problem revolves around the need for robust sentiment analysis models capable of accurately discerning sentiment polarity within diverse datasets. Through the systematic application of CRISP-DM phases - business understanding, data understanding, data preparation, modeling, evaluation, and deployment - the study examines the efficacy of SVM with SMOTE in sentiment classification tasks. The findings demonstrate notable performance metrics, including accuracy (96.50%), precision (95.75%), recall (99.00%), and F-measure (97.34%). The AUC value substantially increases from 0.642 without SMOTE to 0.997 with SMOTE, highlighting its effectiveness in improving sentiment classification accuracy. In addition, The comparative analysis of toxicity values between the first and second videos demonstrates distinct patterns: the first video showcases a Toxicity score of 0.05290, with notable metrics such as Profanity registering at 0.04815. Conversely, the second video exhibits a slightly lower Toxicity score of 0.04744, with varying metrics such as Severe Toxicity at 0.01386.

Penerapan CRISP-DM dalam Klasifikasi Sentimen dan Analisis Perilaku Pembelian Layanan Akomodasi Hotel Berbasis Algoritma Decision Tree (DT) Singgalen, Yerik Afrianto
Jurnal Sistem Komputer dan Informatika (JSON) Vol. 5 No. 2 (2023): Desember 2023
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v5i2.7081

The Cross-Industry Standard Process for Data Mining (CRISP-DM) approach is very relevant in identifying business challenges and producing recommendations in the form of appropriate models to face various business challenges. Sentiment classification is needed to identify and analyze consumer trends and preferences in order to plan risk mitigation strategies related to business sustainability. This study adopts the CRISP-DM method in classifying hotel guest sentiment through review data on the Agoda platform and analyzing sentiment data based on the purchase behavior of related products and services. Meanwhile, the stages in the CRISP-DM method are as follows: the stage of understanding the business context (business understanding), the stage of understanding data characteristics (data understanding), the modeling stage (modeling), the evaluation stage, and the implementation stage (deployment). The results of this study show that ten words are the attention of hotel guests and are dominated by positive sentiment, namely shopping, great, stay staff, clean, location, room, good, mall, and hotel. The classification results using the DT algorithm showed good performance with an accuracy value of 93.91%, a precision value of 90.98%, and a recall value of 97.77%. Â In addition, the AUC value is 0.943 or 94.3%, and the f-measure value is 94.18%. Furthermore, sentiment analysis data can be developed into a Customer Relationship Management (CRM) Â supporting application to analyze guest purchase history data related to sentiment, country of origin, guest type, room type, and length of stay by day, month, and year. Thus, the marketing strategy of hotel accommodation services can be optimized for personalization and increase interest and intention of returning stays.

Implementasi Metode CRISP-DM dalam Analisis Model Pendukung Keputusan Simple Additive Weighting dan Pengembangan Basis Data Riwayat Pembelian Layanan Akomodasi Hotel Singgalen, Yerik Afrianto
Jurnal Sistem Komputer dan Informatika (JSON) Vol. 5 No. 2 (2023): Desember 2023
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v5i2.7153

The development of studies on implementing Simple Additive Weighing (SAW) decision support models in purchasing hotel accommodation services or making stay decisions is limited to hotel recommendations calculated from consumer assessments of criteria with predetermined weights. However, it is necessary to develop a database with interactive visualization of hotel accommodation services and make it easier for consumers to compare and provide ratings. Considering this, this study uses the CRISP-DM method to develop a database based on the purchase history of hotel accommodation services in a business operational area, then uses SAW as a decision support model in the calculation process to produce the best hotel recommendations based on purchase data. The CRISP-DM method consists of business understanding, data understanding, modeling, evaluation, and deployment stages. At the business understanding stage,Â the customer's purchase history data is collected in the Agoda website review column. At the data understanding stage, the data collection process is carried out based on the supporting data of the criteria used. At the modeling stage, the SAW algorithm is used in the calculation process to get the best recommendations. In the Evaluation phase, hotels with the best recommendations are analyzed based on the guest's country of origin, guest sentiment, type of guest staying, room type used, length of stay, and month and year. In the Deployment stage, the database is developed using Oracle Apex and visualized interactively so that system users can understand consumer trends and behavior, especially in making overnight decisions based on purchase history data. Based on data obtained from the Agoda platform, it can be seen that A2 ranks first with a value weight of 0.983, then A3 ranks second with a value weight of 0.982, and A1 ranks third with a value weight of 0.946. Meanwhile, based on data obtained from the Booking.com platform, it can be seen that A3 ranks first with a value weight of 0.983, then A2 ranks second with a value weight of 0.974, and A1 ranks third with a value weight of 0.951. Thus, the SAW decision support model implementation output is not limited to the results of calculations and recommendation tables but includes a database with interactive visuals.

Pemilihan Paket Wisata One Day Tour Menggunakan Model Pendukung Keputusan TOPSIS Singgalen, Yerik Afrianto
Jurnal Sistem Komputer dan Informatika (JSON) Vol. 5 No. 2 (2023): Desember 2023
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v5i2.7231

The development of Labuan Bajo tourism attracts foreign and domestic tourists to explore various natural beauties through fun activities. Tour agents provide a variety of interesting activities and tourist visits to the islands around Labuan Bajo that can provide new experiences for tourists. The activities are sold through One-day tour packages in Labuan Bajo. Still, depending on the service provider, the itinerary description, admission ticket,Â and order of visit to the destination are very complicated. Considering this, this study uses the TOPSIS decision model in choosing One Day Tour tour packages in Labuan Bajo by considering the criteria of price, destination, duration, admission ticket, and service rating. Based on the results of this study, it can be seen that the highest preference value from the ranking results is the Full Day Trip to Explore 6 Destinations in Labuan Bajo and Komodo tour package, with a value of 0.673056111. Furthermore, the tour package that occupies the second position from the ranking results is the 1-day Komodo island Tour hopping around by Speed Boat with a preference value of 0.628303746. Meanwhile, the tour package that occupies the third position from the ranking results is One day Komodo trip with Bintang Komodo Tours with aÂ preference value of 0.53181476. This shows that the TOPSIS method produces recommendations for One Day Tour packages in Labuan Bajo for tourists by considering the price of tour packages, the number of destinations visited, the length of time or duration of tourist time, admission tickets in tour packages, and ratings The services of the travelers were previously related to the tour package. Thus, the selection of TOPSIS-based One Day Tour tour packages can minimize the risk that causes misunderstandings or tourist dissatisfaction related to the tour packages prepared by each travel agent.

Sentiment Classification of Food Influencer Content Reviews using Support Vector Machine Model through CRISP-DM Framework Singgalen, Yerik Afrianto
Jurnal Sistem Komputer dan Informatika (JSON) Vol. 5 No. 3 (2024): Maret 2024
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v5i3.7509

The research problem revolves around the challenges in effectively marketing culinary tourism aligned with tourist preferences in Indonesia, necessitating a substantial exploration of consumer sentiments related to culinary diversity through the lens of food influencer content. Food influencers are crucial in stimulating tourists' interest in gastronomy through culinary tourism in Indonesia. This research reveals challenges in culinary tourism marketing aligned with tourist preferences, necessitating substantial exploration of consumer sentiments related to culinary diversity through food influencer content. The sentiment classification method employed is the Cross-Industry Standard Process for Data Mining (CRISP-DM) using the Support Vector Machine (SVM) algorithm and the SMOTE operator. The data source is derived from a video with the ID PMhfLy_buV8, containing 114,422 comments. This study collects and processes 30,000 comments, resulting in 9,323 data points. The findings highlight the vital performance metrics of SVM models, both with and without SMOTE, showcasing high accuracy, precision, recall, and F-measure values. Specifically, SVM without SMOTE achieves 95.28% accuracy, while SVM with SMOTE achieves 98.67%. Despite some limitations in discerning positive and negative sentiments, indicated by moderate Area Under the Curve (AUC) values (0.608 to 0.658), the overall efficacy of SVM in sentiment analysis for food influencer content is apparent. Drawing from a dataset of 30,000 comments, these insights contribute to advancing sentiment analysis methodologies and offer practical implications for understanding consumer perceptions and behaviors in digital media and influencer marketing. Additionally, the prominence of frequent words such as "bang" (1322), "nonton" (1064), "makan" (921), "yang" (801), "puasa" (711), "tahun" (484), "ngiler" (448), "lagi" (384), "tanboy" (311), and "enak" (315), as extracted from RapidMiner analysis, underscores the significance of language patterns in the realm of food influencer content.

Toxicity Analysis and Sentiment Classification of Wonderland Indonesia by Alffy Rev using Support Vector Machine Singgalen, Yerik Afrianto
Jurnal Sistem Komputer dan Informatika (JSON) Vol. 5 No. 3 (2024): Maret 2024
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v5i3.7563

The music industry's increasing reliance on digital platforms like YouTube for dissemination raises concerns about the potential impact of music videos on viewer sentiment and well-being. This study seeks to assess the toxicity and sentiment of the Wonderland Indonesia music video by Alffy Rev through Support Vector Machine analysis, contributing to our understanding of the effects of music content on online audiences. This research addresses the challenge of sentiment classification in digital content by leveraging the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework. The study aims to enhance sentiment classification accuracy by applying a Support Vector Machine (SVM) with a Synthetic Minority Over-sampling Technique (SMOTE) to address class imbalance issues. The research problem revolves around the need for robust sentiment analysis models capable of accurately discerning sentiment polarity within diverse datasets. Through the systematic application of CRISP-DM phases - business understanding, data understanding, data preparation, modeling, evaluation, and deployment - the study examines the efficacy of SVM with SMOTE in sentiment classification tasks. The findings demonstrate notable performance metrics, including accuracy (96.50%), precision (95.75%), recall (99.00%), and F-measure (97.34%). The AUC value substantially increases from 0.642 without SMOTE to 0.997 with SMOTE, highlighting its effectiveness in improving sentiment classification accuracy. In addition, The comparative analysis of toxicity values between the first and second videos demonstrates distinct patterns: the first video showcases a Toxicity score of 0.05290, with notable metrics such as Profanity registering at 0.04815. Conversely, the second video exhibits a slightly lower Toxicity score of 0.04744, with varying metrics such as Severe Toxicity at 0.01386.

Implementation of Global Vectors for Word Representation (GloVe) Model and Social Network Analysis through Wonderland Indonesia Content Reviews Singgalen, Yerik Afrianto
Jurnal Sistem Komputer dan Informatika (JSON) Vol. 5 No. 3 (2024): Maret 2024
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v5i3.7569

Integrating the Global Vectors for Word Representation (GloVe) Model with Social Network Analysis presents a promising approach for extracting nuanced semantic relationships from Wonderland Indonesia's content reviews. However, the lack of comprehensive studies exploring the effectiveness of this integration, specifically within the context of Wonderland Indonesia's content reviews, necessitates focused research to uncover its potential impact and applications. This study investigates the reception and impact of the "Wonderland Indonesia" video content by Alffy Rev ft. Novia Bachmid (Chapter 1) within the YouTube community using a comprehensive methodology based on CRoss-Industry Standard Process for Data Mining (CRISP-DM), topic analysis, and Social Network Analysis (SNA). Through topic analysis, the content's main themes and narrative elements were identified, shedding light on its storytelling effectiveness. Furthermore, sentiment analysis using Vader was conducted on 2204 out of 24185 posts, revealing that 1369 (92%) exhibited positive sentiment, 427 (31.19%) had neutral sentiment, and 850 (62.09%) contained negative sentiment. Additionally, sentiment analysis using TextBlob was performed on the same subset of posts, with 1369 (40) posts exhibiting positive sentiment, 599 (43.75%) with neutral sentiment, and 730 (53.32%) expressing negative sentiment. Notably, metrics such as toxicity (highest value: 0.90780) and severe toxicity (highest value: 0.95021) exhibited varying prominence within the analyzed content. These findings enable targeted interventions and content moderation strategies to promote healthier online discourse. The SNA uncovered intricate social dynamics and interaction patterns among viewers, emphasizing the video's ability to foster engagement and community interaction. This study underscores the significance of creative storytelling and community engagement strategies in digital content creation, with implications for audience participation and community development within the digital sphere. Future research could explore the longitudinal effects of such content strategies on audience retention and community engagement.

Title

Found 9 Documents
Search
Journal : Jurnal Sistem Komputer dan Informatika (JSON)

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 9 Documents Search Journal : Jurnal Sistem Komputer dan Informatika (JSON)

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 9 Documents
Search
Journal : Jurnal Sistem Komputer dan Informatika (JSON)