Sentiment analysis is a review analysis process used to determine whether an opinion is neutral, negative, or positive. Sentiment analysis can be done using lexicon-based or machine learning-based approaches. Lexicon can perform sentiment analysis without training data because it is dictionarybased but performs worse than machine learning. Machine learning can perform well in completing sentiment analysis but requires training data so that the model does not experience underfitting. In the case of sentiment analysis on big data, manual labeling of training data is an inefficient job. Support vector machine (SVM) has the opportunity to be used together with the active learning (AL) method to make small training data but still have good performance. This research proposed a hybrid lexicon and AL-SVM method to complete sentiment analysis on big data tourism. This research used polarity from the valence aware dictionary and sentiment reasoner (VADER) lexicon as a reference for the query by user process from the ALSVM to automate the sentiment analysis process on big data. The experimental results showed that using the hybrid lexicon and AL-SVM increased the sentiment analysis performance compared to the VADER lexicon, SVM, and lexicon SVM, which run separately.

Bali is a tourist destination with the largest market share in Indonesia. So the progress of Bali tourism has a major contribution to the country's foreign exchange earnings. This of course cannot be separated from the support of the Bali hotel industry. This research intends to reveal the potential for hotel development in Bali based on sentiment analysis and analysis of tourist characteristics through data analytic processes. From the results of the sentiment analysis, several aspects were obtained that can be used as references to strengthen and maintain the positive image, such as the friendliness of the staff and improving aspects that cause tourist dissatisfaction, such as room replacement. So hotel services in Bali will get better. The potential for hotel improvement in Bali can also be supported by understanding the characteristics of the guests. This research uses a clustering method to group tourists who stay in star hotels and budget hotels. Understanding the characteristics of each tourist cluster can be a reference for marketing strategies and decision-making in this industry.

Lontar is one of Bali's cultural heritages that must be preserved. In recent years, the lontar conservation team from Gianyar Regency found that many lontar collections owned by residents were damaged. This damage is caused by various factors, one of which is the lack of knowledge of residents in caring for Lontar. Not only residents' collections but some damaged collections were also found in the Bali Provincial Cultural Office. As a solution, this research develops a lontar digitization system that can convert Latin text into Balinese script using the Jaro Winkler Distance method. The result of the development of this lontar digitization system is that the system can produce output that has content that is close to the original lontar. Of the 10 sheets tested, an accuracy rate of 92.7% was obtained. In media expert testing by assessed the writing structure by Balinese script rules, which obtained an accuracy rate of 93.6%. In the accuracy test involving the spelling checker process using the Jaro Winkler distance method, the accuracy was 93.8%.

With the increasing need for digital images in everyday life, images are collected through various devices such as digital cameras, cell phone cameras, and scanners. This image data will be further processed, one of which is to segment objects from the background. The technique that can be used is segmentation using the LAB color space. This technique is done by converting the image color space into LAB color space so that the object or foreground can be sepa rated from the background. This research uses 20 random images from 3 sources: The Oxford-IIIT Pet dataset, Github Real Python material, and DeepLontar da-taset. The experimental results show that The Oxford-IIIT Pet dataset and Github Real Python material have a more extended range of minimum-maximum values of L, a*, and b* components compared to DeepLontar dataset. This extended minimum-maximum value range causes the object images in The Oxford-IIIT Pet dataset and Github Real Python materials to be more visually visible (segmented) than in the DeepLontar dataset.

Flowers play a crucial role as elements in Hindu ceremonies in Bali. In the Balinese Hindu tradition, flowers serve as ritual instruments that carry spiritual significance and profound symbolism. Many of the younger generation in Bali today struggle to identify various types of flowers used in Balinese Hindu ceremonies. The utilization of machine learning technology can provide a solution to overcome the limited knowledge faced by the younger generation in Bali when identifying flower types for Balinese Hindu ceremonies. Convolutional Neural Network (CNN) is one highly effective machine learning method for image classification. One efficient CNN model for identification is the ResNet152 architecture. In this study, image data consists of flowers used in Balinese Hindu ceremonies, encompassing five types of flowers. The test results demonstrate an accuracy rate of 98%, with precision at 94%, and recall at 93%. The detection outcomes of flowers used in Balinese Hindu ceremonies through the Convolutional Neural Network (CNN) method yield a satisfactory accuracy value.

The use of Deep Learning technology, especially image recognition, is an appropriate for classifying digital images. This research aims to classify images of typical Balinese snacks by identifying the optimal model within the VGG-16 architecture. The evaluation involves comparing accuracy, recall, precision, and f1- score across different test scenarios, including dropout weight, batch sizes, and epochs. The training dataset for this research comprises 2,445 image data points of typical Balinese snacks, distributed among 10 classes. These classes include Klepon (320 images), aLaklak (207 images), Kaliadrem (222 images), Jaje Lukis (327 images), Jaje Batun Bedil (189 images), Pisang Rai (200 images), Jaje Piling-piling (234 images), Jaje Wajik (241 images), Ongol-ongol (308 images), and Bubur Injin (197 images). Each class is represented by 50 image datas, resulting in a total of 500 image datas per class. The best model derived from the VGG16 architecture for classifying typical Balinese snacks achieved an accuracy level of 97.5%, precision of 87.9%, recall of 87%, and an f1-score of 87.4%. This performance was achieved with a dropout test parameter of 20%, a batch size of 64, and an epoch setting of 1000. The evaluation was conducted on test image data separate from the training and validation datasets.

Indonesia boasts a unique traditional woven fabric called ikat gringsing, created using the double ikat technique and natural materials from Tenganan Village, Pegringsingan, Bali. The motifs are inspired by the surrounding nature, featuring animals and plants commonly found in the area. The intricate and diverse designs of gringsing woven cloth have attracted many tourists, and this cloth has the potential to boost Bali's and Indonesia's tourism sector. However, it can be challenging for tourists to find detailed information about gringsing woven fabrics, and some may not even know the names of the motifs that catch their eye. To address this issue, a deep learning model utilizing CNN has been developed to introduce ikat gringsing woven fabric motifs. With an accuracy rate of 82%, the model successfully recognized 402 fabric motifs out of a total of 470 datasets. Utilizing the dropout technique to prevent overfitting resulted in consistent accuracy values throughout the training and testing process. The model's validation results in the training process averaged at 86%, while the testing process yielded an accuracy rate of 82% out of the 10 folds.

Tourism development is currently very rapid due to advances in information technology. The growth of various digital tourism platforms makes it easy to obtain tourism information. Almost every time tourists leave traces of their travel experiences on digital platforms, one of them is a review. Until now, the review is only greatly beneficial to prospective tourists in terms of supporting decision-making about the consumption of tourism products and services to minimize the risk of failure. This research analyzed large amounts of reviews, known as big data, to benefit stakeholders. This research aimed to describe the image of Bali tourism that originated from tourist attraction reviews on TripAdvisor. Sentiment analysis was carried out using the Vader Lexicon method to obtain the image clarified using term frequency, bigrams, and topic-based trigrams. The analysis obtained several positive images of tourist attractions in Bali, including beautiful beaches, amazing temples, and friendly locals. Meanwhile, we recommend improvements to several negative images that we found such as dirty tourist objects, plastic waste, and disturbances from hawkers in order to get a better image of Bali tourism.

This research aims to visually predict activities in the Tri Pramana Learning Concept in a virtual reality (VR) environment using the ResNet-50 deep learning architecture. The method in this research consists of dataset preparation (data acquisition, frame extraction, data cleaning, image resizing, data subsetting), ResNet-50 model building, and evaluation. The data used in this study comes from learning recordings in a virtual classroom environment of fiber optic splicing practicum. The total number of images in this dataset is 2,163 which is divided into training subset (70%), validation subset (20%), and testing subset (10%). This research focuses on experimenting with epoch variations of 100, 200, 300, and 400 to produce the best model. Through the investigation, it was found that the model with epoch 400 was able to provide the best performance with Accuracy 97.72%, Precision 97.81%, Recall 97.77%, dan F1- Score 97.79%. Future experiments will focus on the variation of learning rate and batch hyperparameters as well as comparisons with other deep learning architectures to predict activities in the Tri Pramana Learning Concept in virtual environments.

Musik gamelan angklung Bali lewat gelombang bunyi yang dihasilkannya mampu menginterferensi gelombang pikiran manusia untuk menurunkan frekuensi gelombang yang dipancarkan oleh otak. Tujuannya untuk mempengaruhi kondisi psikologi yang berkaitan dengan suasana hati agar mengarah pada tingkat stress positif dengan tingkat energi rendah maupun tinggi. Musik dengan tingkat stress positif dan tingkat energi rendah masuk ke dalam kategori suasana hati tenang atau contentment, jika tingkat stress positif dan tingkat energi tinggi masuk ke dalam kategori suasana hati senang atau exuberance. MIR (Music Information Retrieval) adalah bagian dari Data Mining yang menggali informasi mengenai data musik, salah satunya yaitu klasifikasi suasana hati yang diinterpretasikan oleh potongan data musik. Penelitian ini merancang dan membangun sistem klasifikasi untuk mendeteksi suasana hati musik gamelan angklung Bali menggunakan algoritma K-NN dan K-NN berbasis Algoritma Genetika. K-NN dapat mengatasi masalah klasifikasi dengan baik, namun dibalik keunggulannya,

Penelitian bertujuan untuk mengetahui sejauh mana suatu aplikasi Sistem Informasi Wisata di Pulau Ambon dapat diterima masyarakat dan dan seberapa besar tingkat kepuasan masyarakat akan adanya Sistem Informasi yang dibuat. Penulis dalam hal ini membuat suatu aplikasi untuk melihat persepsi pengguna terhadap aplikasi sistem informasi pariwisata di Kota Ambon berbasis Android dengan mengandalkan fitur yang dapat diakses dengan mudah dan tampilan yang sederhana, Selanjutnya untuk melihat sejauh mana tingkat kebergunaan aplikasi, dilakukan pengujian usability dengan melibatkan pengguna, bagaimana pengguna menggunakan sistem serta permasalahan yang dihadapi. Pengujian dilakukan menggunakan kuisioner untuk mengukur kepuasan pengguna dan untuk mengetahui opini pengguna terhadap aplikasi yang digunakan. Metode yang digunakan adalah metode Waterfall, yang terdiri dari tahapan Analisis Kebutuhan, Perancangan, Implementasi dan Pengujian. Pengumpulan data dilakukan dengan menyebar kuesioner tertulis kepada 100 orang responden di beberapa spot daerah wisata. Data kuesioner yang telah diberikan kepada responden dianalisa menggunakan model skala likert. Dari diagram hasil dapat dianalisa sebagian besar responden setuju bahwa aplikasi mampu memenuhi unsur usability yakni : learnability, eficiency, memorability, errors, dan satisfaction. Sehingga dapat diambil kesimpulan bahwa aplikasi sistem informasi wisata Pulau Ambon berbasis Android mampu memberikan informasi tempat wisata, kegiatan wisata dan daftar kuliner yang bisa digunakan untuk lebih memudahkan wisatawan dalam merencanakan tujuan wisata yang diinginkan.

One of the applications of computer vision in the popular culture is food recognition which is popularized in the internet with a “hotdog and not hotdog” problem. Food recognition is also useful in many popular lifestyle apps such as calorie counter app or any diet related app. In this paper is proposed a CNN aided technique for recognizing food that is common in Indonesia. The technique is consist of 3 main phases, one is pre-processing normalize the data one is model formation and training which is known as the common binary classifier template, boosted with pooling and evaluated by cross-entropy technique in this paper the model used is the pre-trained model to test the testing data afterward, which is show a promising result with a relatively short training time. The experiments focused on how CNN can be used as a component to recognize food so that in the future it can be used to develop better calorie counter applications. In this experiment 10,000 data were used for training and 50 tests for each food category with a total of 500 food image data used for testing with the best accuracy reaching 88% for one of the categories.

Dynamic price in the hospitality industry is a term to indicate the selling prices of hotel rooms that can be changed at certain times dynamically. Based on the author’s observation at Mercure Bali Nusa Dua Hotel, the process of deciding the dynamic price level in the hotel is greatly influenced by individuals hence highly subjective, therefore prone to inconsistent pricing mechanism and decision which in turn affects the revenue of the hotel. In this paper, the authors propose expert systems combined with classical probabilities to solve the problem. Fuzzy logic and forward chaining are used to form inference rules that produce a knowledge base of the system. Next, the classical probability is used to calculate the confidence level of the conclusion. The algorithms are tested with the price level of the hotel in 2018. The result shows an initial accuracy of 73.66% with average deviations of 0.36. By classifying the deviations with the rule-based classifier method, the accuracy increases to 90.77%. It is shown that the difference between the actual data is small. The proposed technique potentially increases the hotel’s revenue. The usability score of the proposed system is 91.88, indicating the usability of the proposed system is grade A and excellent rating. 

Gringsing merupakan salah satu kain tradisional yang menjadi ciri khas dari Desa Tenganan Pegringsingan. Gringsing terbilang sangat unik karena pada pembuatannya menggunakan teknik dobel ikat, dimana teknik ini hanya bisa ditemukan di tiga tempat saja di dunia, yaitu Jepang, India dan Indonesia teptnya di Desa Tenganan Pegringsingan. Pada tahun 2016 kain gringsing mendapat sertifikasi Indikasi Geografis dari Kementerian Hukum dan Hak Asasi Manusia Republik Indonesia. Penelitian ini bertujuan untuk membangun sebuah model deep learning untuk mengenali motif-motif kain tenun gringsing dan mengetahui performa model deep learning yang dibangun. Sehingga masyarakat dapat lebih mudah untuk mengenali motif motif kain tenun gringsing tanpa harus memiliki kemampuan khusus. Model deep learning pengenalan pola motif kain tenun gringsing dibangun menggunakan metode Convolutional Neural Network (CNN) dengan model arsitektur AlexNet. Pengujian dilakukan untuk mengetahui performa model yaitu waktu training, akurasi, presisi, recall dan nilai f-measure. Berdasarkan hasil pengujian model yang dibangun mampu menyelesaikan training 100 epoch dengan waktu 19,33 Jam, serta memiliki nilai akurasi sebesar 76%, presisi 74,1%, recall 72,3%, dan F-measure sebesar 0,73.

Pupuh is a basic of tembang for someone that can later be used for further learning to a higher level that is Sekar Madya and Sekar Agung. Balinese culture is an important feature to determine the ethnicity, identity of a group. Then it should be the Balinese using and preserve Pupuh that has been inherited by the ancestors. The media for Balinese pupuh learning will be use an Android Device with MFCC method. To attract the interest of young peoples to learn Balinese pupuh, then this Balinese pupuh learning system based-Android was created. This application will use the sound recorded by an Android devices and sent to the Python server for systematic calculations, then returned to the Android devices and will get the correct or wrong answer which according to server calculation. By developed this Balinese pupuh learning system, expected can help the teachers or Balinese language teachers to teach their students about Balinese pupuh on a mobile and practical.

Teknologi   merupakan   salah   satu   aspekyang   tidak   bisa   dipisahkan,   karena   dengan   adanya   teknologi mempermudah  komunikasi.  Dunia  industri  pariwisata  tidak  bisa  lepas  dari  peran  teknologi.  Dari  penelitian  ini diharapkan  wisatawan  baik  lokal  maupun   mancanegara  memberikan  persepsinya  terhadap  efektivitas  e-Kios sehingga ke depan dapat menjadi sumber informasi yang handal tentang pariwisata Bali.Alat  ukur  yang  digunakan dalam    dalam  penelitian  ini  adalah  menggunakan  skala  likert  dari  rensis    guna  mendapatkan  pesepsi  total  dari efektivitas e-Kios pariwisata Bali.Dalam  penelitian  didapat  adanya  suatu  fakta  bahwa  e-Kios  pariwisata  Bali  belum  mencapai  hasil  maksimal dengan  kata  lain  kurang  efektif  dikarenakan  belum  secara  murni  online  sehingga  informasi  e-Kios  tidak up-date. Meskipun  demikian  e-Kios  mendapat  respon  positif  karena  menambah  pencitraan  Bali  sebagai  destinasi  favorit. Seperti  halnya  informasi  spiritual  dan  sosial  budaya  Bali  tetap  menjadi  informasi   yang  paling  dibutuhkan wisatawan.  Harapan  kedepan  informasi  e-Kios  lebih  dimaksimalkan  termasuk  segi  brainware,  hardware  dan software.

Game is one kind of attractions that exist in local rural communities. In Bali, traditionally, there are various types of games played by children and adults. The Balinese traditional games such as tajog, gangsing, petaumpet, etc. can be found in particular villages. Not only games, Bali also has various kinds of folklores or legends, known for generations, namely Jayaprana, I Rajapala, etc. The folklores and legends had not been much developed in the form of ICT-based game. In this case, a very interesting folklore was told and developed in the form of a game. The aim of the game development was to introduce and preserve Balinese local wisdom. The folklore of Balinese culture that was developed was “I Rajapala as a Hunter”. The method used in developing the game was a method of waterfall. Limited test results showed that it was technically able to be run as the game designs: (1) There are three missions of the game: shooting 3 pigs or birds, finding a place to rest, and hiding one of the angel' s shawls; (2) The difficulty level of the game consists of three levels: easy, medium, and hard to complete the three missions of the game. The results of limited application tests showed that 93.4% (3 stars) of players could complete the game mission, while the remaining 6.6% (2 stars) failed to complete the 3 game missions. The benefit gained by the players is knowing the story of I Rajapala as a Balinese cultural heritage.

Kawi Language is a kind of language which has developed in Java during the Hindu-Buddhist era in the ancient Indonesian archipelago. As the time passed, Kawi began to come to extinction. To overcome this, preservation of Kawi language can be done by enhancing the understanding of culture through the use of technology. Technology that can be used for cultural preservation is one of NLP based applications such as Kawi Language to Bahasa Translator Application which is developed in this research. The translator application is an application that translates from one language to another. The result of translation of the translator application is a language that is easy to understand by the people or commonly used language. This translator application is developed to processes Kawi Language text into bahasa based on search method which each word inserted. The stemming method used is Bobby Nazief & Mima Adriani Algorithm using Kawi Language rule. The translated Kawi Language text can be a word, sentence or paragraph with the result of the Text in Bahasa as well as words, sentences or paragraphs.

Sejarah masa lampau merupakan warisan dari para pendahulu yang seharusnya dijaga dan dilestarikan oleh para generasi muda berikutnya. Salah satunya adalah mahakarya yang terkenal dari Indonesia yaitu wayang purwa. Wayang purwa disebut juga wayang kulit karena terbuat dari kulit lembu. Wayang purwa sendiri biasanya menggunakan cerita Mahabharata dan Ramayana. Salah satu solusi yang dapat digunakan adalah mengimplementasikan wayang purwa dalam bentuk aplikasi media pembelajaran berbasis mobile. Aplikasi media pembelajaran dibangun untuk dapat dijalankan pada smartphone berbasis Android menggunakan software Unity yang mempunyai fitur 2D dengan user interface, drag and drop dan scripting bahasa pemrograman C#. Aplikasi media pembelajaran wayang purwa memberikan pengetahuan tentang rupa dan karakter wayang purwa dalam bentuk game puzzle dan juga kuis. Berdasarkan hasil pengujian usability testing, memanfaatkan kuisioner terhadap 30 responden diperoleh hasil pada pengujian aspek grafis visual sebanyak 57% responden memberikan respon baik sedangkan pada pengujian aspek entertaintment dan pembelajaran sebanyak 90% memberikan respon baik. Selain itu berdasarkan pengujian dengan menggunakan metode blackbox keseluruhan fungsionalitas aplikasi telah berjalan dengan baik. Dengan adanya aplikasi media pembelajaran ini diharapkan dapat meningkatkan minat para generasi muda untuk mempelajari wayang purwa.

Introduction, The Balinese writing is unique in its almost identical form, and some writings are distinguished by a single line stroke. The Balinese writing is complex, in the form of a combination of several characters in a syllable group with the position of surrounding the main character in Balinese script called the sound usage. The main characters generally have a combination of follower characters in front, behind, above and / or under the main characters; even in each position can contain combinations of more than one character, thus forming a model that is much more complex than the Latin script. Methodology: DBSCAN (DensityBased Spatial Clustering of Application with Noise) algorithm is suitable for clustering process. DBSCAN has an algorithm that builds high-density areas into clusters and finds clusters of any kind in a spatial database containing noise inside. The clustering process is as an early stage in Balinese Optical Character recognition (OCR) System on Kakawin Books into poetry in Latin Letters. Trials using a sample image of Balinese script are taken from Kakawin Ramayana Book. The process begins with binary, followed by cropping automatically to get rows per line of writing. After that they are processed with Clustering Process to get the character objects. Variations in minimum point value (minpts) and epsilon (eps) values. The results obtained by DBSCAN Algorithm with the minimum value of points 2 and 3, epsilon = sqrt (2) and sqrt (3) succeeded in clustering with error percentage below 3%. Conclusion: DBSCAN algorithm is very good for conducting Complex Balinese Handwriting Process.

Lontar is one of the cultural heritage which has information about history of Balinese civilization in the past away. Problems encountered today are that the lontar are not well maintained. While, the lontar that used as letter of Balinese alphabet will be worn out soon, because it doesn’t have good endurance for long time. The exploiting of technology is one of media that can be used as a solution to solve the problems. The digitalization process of letter Balinese alphabet can be done by rewrite the script of that lontar in Balinese alphabet by using translation script with Rule Base and Levenshtein Distance Approach. The exploiting of technology will make the lontar becoming digital form and it won’t be worn out when the lontar is kept safe for long time and information that consisted in lontar can be protected for long time.

Balinese people have one of the civilization histories and cultural heritage are handwritten in Balinese script on palm leaves known as Balinese Papyrus (LontarAksara Bali). Until now that cultural heritage is still continuously strived its preservation along with the implementation begin to be abandoned in public life. Some of Balinese Papyrus now begins to rot and fade under influenced by age. Information technology utilization can be a tool to solve the problems faced in the preservation of the Balinese papyrus. By using digital image processing techniques, the papyrus script can be reconstructed digitally so that it can be retrieved and store the content in the digital media. Balinese papyrus reconstructed through several processes from scanning into a digital image, performing preprocessing for image quality improvement, segmenting the Balinese characters on image, doing character recognition using LDA algorithm, rearranging the result of recognition in accordance with the original content in papyrus, and translating that characters result into Latin. LDA algorithm quite successfully performs the classification associated with handwritten character recognition.

The Balinese papyrus is one of the media to write the ideas from minstrels in ancient times. Currently, many of ancient literature that written in papyrus very difficult to identify because the writings were beginning to rot or fade influenced by age. This study takes the Balinese papyrus characters as its object. The improvement of the image quality in image processing of the characters here refers to the change for the quality of the image before its next development. There is the process of identifying the color space to spot the shape of the papyrus characters. The difference of the colors is so small so that it needs CIELab to process the identification out of its background and noise.

In the ancient Balinese Civillization, lontar is a way of Balinese people to preserving information. Lontar as a medium of writing, still taught today as a way to preserving Balinese heritage among Balinese youngster, but Lontar is struggling to survive in modern life of Balinese people ever since. One of the problem is, mainly that the lontar are not well maintained lontar also a medium of writing also found to be worn out soon so its not a good medium of storing information, lontar itself it doesn’t have good endurance properties against times and natural weather. Author propose to digitize the experience of writing lontar with the use of software. The digitalization process of each letter in Balinese alphabet can be done by rewriting the script of that lontar in Balinese alphabet by using translation script powered by Rule Base and assisted with Levenshtein Distance Approach. The touch of Digital technology will preserving the lontar into a digital format thus making the problem of worn out lontar becoming obsolete making the information that written in this digital lontar protected for a long time.

Papyrus script is a cultural heritage in Bali. As we know, that the papyrus is a cultural matter which is rich in valuable cultural values. Issues or a problem encountered today is that the papyruses are not well maintained. Thus, many papyruses become damaged because it is not stored properly. Papyrus script was written using Balinese script’s characters which having different features compared with Latin’s characters. Balinese script can be recognized with feature extraction owned by each Balinese script. KNN is a classification algorithm based on nearest neighborhood. KNN can be used to classify Balinese script’s features so that the test Balinese script’s features which having nearest neighborhood value with the trained Balinese script’s features will be recognized as the same Balinese script.

One of the cultural heritages which are now starting to be forgotten is the papyrus library. Balinese papyrus is one of the media to write the ideas from minstrels in ancient times. Currently, many of ancient literature that written in papyrus very difficult to identify because the writings were beginning to rot or fade influenced by age. The introduction of Balinese scripts on papyrus can be done first by performing papyrus digitalization. The papyruses are scanned to become one image file. Further the papyrus image is done by Thresholding since prior to the thinning process it is required a binary image. The application of Thinning Zhang-Suen method is very effective because from the original image with 2 sub iteration until yielding in 1 pixel. The benefit of this research is to improve the quality of the image and further segmented to read papyrus making it easier to read text on the papyrus.

Ketersediaan dokumen teks bahasa Bali yang meningkat jumlahnya membuat proses pencarian informasi pada dokumen teks berbahasa Bali menjadi semakin sulit. Mengklasifikasikanya secara manual menjadi tidak efisien mengingat peningkatan jumlah dokumen yang semakin banyak. Pada penelitian ini dikembangkan sebuah aplikasi yang dapat mengklasifikasikan teks bahasa Bali ke dalam kategori yang ditentukan. Aplikasi ini menggunakan metode Naive Bayes Classifier (NBC) dan metode Information Gain(IG) untuk seleksi fitur. Aplikasi ini diuji dengan teknik cross validation. Hasilnya adalah nilai rata-rata akurasi dari 10 fold cross validation sebesar 95,22%.

Balinesse lontar digitalization generates image file which acquired through a scanner or camera. Lontar image has noise because the results of the acquisition of the original lontar contained brown color that exist on the leaves. Therefore this paper focuses on improving the quality of the image to remove noise contained in the image by thresholding process. The method used in this paper is a Local Adaptive Thresholding. The test results in this paper generates the best image with the window (W)=70 and the threshold value (C)=0.05 which proved to remove noise at most of the few testing that has been done in this paper.

Balinese papyrus is one of the media to write the ideas from minstrels in ancient times. Currently, many ancient literatures which are written in the papyrus, that also very difficult to identify because the writings were beginning to rot or fade influenced by age. In this study takes the Balinese papyrus characters as its object. The improvement of the image quality in image processing of the characters here refers to the change for the quality of the image before its next development. There is the process of identifying the color space to spot the shape of the papyrus characters. The difference of the colors is so small so that it needs CIELab to process the identification out of its background and noise.

Digitalisasi lontar menghasilkan file-file citra yang diakuisisi melalui scanner ataupun kamera. Citra lontar tentu saja memiliki noise karena hasil akuisisi pada lontar aslinya terdapat warna kecoklatan yang ada pada daun tersebut. Maka dari itu paper ini berfokus pada peningkatan kualitas citra untuk menghapus noise yang ada pada citra dengan melakukan proses Thresholding. Metode yang digunakan dalam paper ini adalah Local Adaptive Thresholding. Hasil uji coba pada paper ini menghasilkan citra terbaik dengan window (W) = 70 dan nilai ambang (C) = 0.05 yang terbukti menghapus noise paling banyak diantara beberapa uji coba yang telah dilakukan pada paper ini

Feature extraction is an important process in character recognition system. The purpose of this process is to obtain special feature from a character image. This paper is focuses on how to obtain special feature from a handwritten Balinese character image using zoning. This algorithm dividing Balinese character image into multiple regions, then a special feature on each region resulting the data extracted feature. The test result in this paper generates a various semantic and direction feature data. This is because this paper using handwritten Balinese character. Furthermore, the features that produced in this paper can be used on Balinese character image recognition process.

Papyrus script is a cultural heritage in Bali. As we know, that the papyrus is a cultural matter which is rich in valuable cultural values. Issues or problems encountered today is that the papyrus are not well maintained. Thus, many papyrus becomes damaged because it is not stored properly. Papyrus script was written using Balinese script’s characters which having different features compared with Latin’s characters. Balinese script can be recognized with feature extraction owned by each Balinese script. KNN is a classification algorithm based on nearest neighborhood. KNN can be used to classify Balinese script’s features so that the test Balinese script’s features which having nearest neighborhood value with the trained Balinese script’s features will be recognized as the same Balinese script.

One of the cultural heritage which are now starting to be forgotten is the papyrus library. Balinese papyrus is one of the media to write the ideas from minstrels in ancient times. Currently, many of ancient literature that written in papyrus very difficult to identify because the writings were beginning to rot or fade influenced by age. The introduction of Balinese scripts on papyrus can be done first by performing papyrus digitalization. The papyruses are scanned to become one image file. Further the papyrus image is done by Thresholding since prior to the thinning process it is required a binary image. The application of Thinning Zhang-Suen method is very effective because from the original image with 2 sub iteration until yielding in 1 pixel. The benefits of this research is to improve the quality of the image and further segmented