Presence using face already widely adopted as a way of monitoring employee attendance. Research on using facial Presence never been done before by applying algorithms and algorithms Eigenface linear discriminant analysis (LDA). However, previous research has found that there are still weaknesses in the algorithms used. The weakness is that the process of identifying which takes a long time because the process of calculating the value carried on the overall image or image and the distance of the face of the webcam can affect the process of identifying faces. In this study, the algorithm used is haar cascade classifier algorithm. Haar classifier cascade or known by other names haar-like features are the rectangular features (square function), which gives an indication of the specifics on a picture or image. Principle Haar-like features are recognizing objects based on simple values of the features but not the pixel values of the object image. This method has the advantage that the computation is very fast, because it depends on the number of pixels in a square instead of each pixel value of an image. Haar classifier cascade also still be able to identify faces even if the distance face with the webcam is considerably due to the value of the facial features can still be identified. Results from this study that the system can identify the face with a good degree of accuracy. Tests carried out to 13 employees Starcross Store with each employee doing 30 times the experiment presence. Attendance successful has the success rate is 87% and 13% of the total failure of the experiment 390 times. Some absences failed to happen because there are several factors that can affect attendance as high luminance, uplifted head position, and the use of attributes (hats, glasses, etc.).Keywords : Presence, face recognition, Haar cascade classifier algorithmPresensi menggunakan wajah sudah banyak diterapkan sebagai cara untuk pemantauan kehadiran pegawai. Penelitian tentang presensi menggunakan wajah pernah dilakukan sebelumnya dengan menerapkan algoritma eigenface dan algoritma linear discriminant analysis (LDA). Namun dari penelitian sebelumnya telah ditemukan kelemahan yaitu pada proses pengidentifikasian yang membutuhkan waktu cukup lama dikarenakan proses perhitungan nilai dilakukan pada keseluruhan citra atau image dan jauhnya jarak wajah dari webcam dapat mempengaruhi proses pengidentifikasian wajah tersebut. Pada penelitian ini algoritma yang digunakan adalah algoritma haar cascade classifier. Haar cascade classifier atau yang dikenal dengan nama lain haar-like features merupakan rectangular features (fungsi persegi), yang memberikan indikasi secara spesifik pada sebuah gambar atau image. Prinsip Haar-like features adalah mengenali obyek berdasarkan nilai sederhana dari fitur tetapi bukan merupakan nilai piksel dari image obyek tersebut. Metode ini memiliki kelebihan yaitu komputasinya sangat cepat, karena hanya bergantung pada jumlah piksel dalam persegi bukan setiap nilai piksel dari sebuah image. Haar cascade classifier juga masih dapat mengidentifikasi wajah walaupun jarak wajah dengan webcam terbilang jauh dikarenakan nilai fitur wajah masih dapat diidentifikasi. Hasil dari penelitian ini bahwa sistem dapat mengidentifikasi wajah dengan tingkat akurasi baik. Pengujian dilakukan kepada 13 karyawan Starcross Store dengan masing-masing karyawan melakukan 30 kali percobaan presensi. Absensi yang berhasil memiliki nilai keberhasilan 87% dan 13% gagal dari total percobaan 390 kali. Beberapa absensi yang gagal terjadi karena ada beberapa faktor yang dapat mempengaruhi absensi seperti pencahayaan yang tinggi, posisi kepala yang mendongkak dan penggunaan atribut (topi, kacamata, dsb).Kata Kunci : Presensi, Pengenalan Wajah, Algoritma Haar Cascade Classifier

Abstrak - Penelitian untuk mencari model prediksi curah hujan yang akurat di berbagai bidang sudah banyak dilakukan, maka dilakukan di-review kembali guna membantu proses penyaliran dalam perusahaan tambang. Review dilakukan dengan membandingkan hasil dari setiap model yang telah dilakukan pada beberapa penelitian sebelumnya. Penelitian ini menggunakan metode kuantitatif. Model yang dibandingkan pada penelitian di antaranya yaitu model Fuzzy, Fast Fourier Transformation (FFT), Emotional Artificial Neural Network (EANN), Artificial Neural Network (ANN), Adaptive Ensemble Empirical Mode Decomposition-Artificial Neural Network (AEEMD-ANN), E-SVR-Artificial Neural Network (E-SVR-ANN), Artificial Neural Network Backpropagation (BPNN), Adaptive Splines Threshold (ASTAR), Seasonal First-Order Autoregressive (SAR), Gumbel, Autoregressive Integrated Moving Average (ARIMA), Feed Forward Neural Network (FFNN), Support Vector Machine (SVM), Adaptive Neuro-Fuzzy Inference System (ANFIS), dan Artificial Neural Network-Fuzzy (ANN-Fuzzy). Hasil dari review menyimpulkan bahwa model Artificial Neural Network memiliki beberapa kelebihan dibandingkan dengan metode yang lain, yakni ANN mampu memberikan hasil yang dapat mengenali pola-pola dengan baik dan mudah dikembangkan menjadi bermacam-macam variasi sesuai dengan permasalahan maupun parameter yang ada, sehingga ANN direkomendasikan untuk perhitungan prediksi hujan. Abstract - Various kinds of research have been carried out to find accurate models to predict rainfall in various fields, so the research that has been done previously was reviewed again to help the drainage process in mining companies. The review is done by comparing the results of each model that has been conducted in several previous studies. This research used quantitative methods. Models compared in this study include the Fuzzy model, Fast Fourier Transformation (FFT), Emotional Artificial Neural Network (EANN), Artificial Neural Network (ANN), Adaptive Ensemble Empirical Mode Decomposition-Artificial Neural Network (AEEMD-ANN), E-SVR -Artificial Neural Network (E-SVR-ANN), Artificial Neural Network Backpropagation (BPNN), Adaptive Splines Threshold (ASTAR), Seasonal First-Order Autoregressive (SAR), Gumbel, Autoregressive Integrated Moving Average (ARIMA), Feed Forward Neural Network (FFNN), Support Vector Machine (SVM), Adaptive Neuro-Fuzzy Inference System (ANFIS), and Artificial Neural Network-Fuzzy (ANN-Fuzzy). The results of the review concluded that the Artificial Neural Network model has several advantages compared to other methods, namely ANN is able to provide results that can recognize patterns well and easily developed into a variety of variations in accordance with existing problems and parameters, so ANN is recommended for rain prediction calculation.

No abstract

There Improvements in the local economy, especially for the poor through open and sustainable tourism management, are believed to be achieved through the empowerment of the tourism sector. Mapping the potential of village tourism in the Triharjo village area is one of the essential things. Identification and mapping of village tourism potential needed in order to implementation community-based tourism (CBT). This research aims to identifying and mapping the potential of village tourism in order to produce a profile of village tourism potential and identify opportunities for developing village tourism potential. The object of this study is Triharjo village, Pandak District, Bantul Regency, Yogyakarta. This research was conducted with a qualitative approach. Collecting data in this study used several research instruments, such as in-depth interviews, focus group discussions (FGD), observations, and document studies. Based on research finding while the communities and local governments of Triharjo village recognize that not all village tourism potentials are well managed. The results of the mapping of village tourism potential provide them that the involvement of local communities in the planning and management of a village tourism potential is needed and have a positive impact on the longterm. The empowerment of the local economy, especially the poor, is believed to be achieved through the empowerment of the tourism sector. Community-based tourism emphasizes community ownership and active participation, provides education to local communities, promotes and protection of culture and the environment.

Human beings as creatures of God has tremendous advantages in terms of process knowledge . Knowledge , ideas , predictions , and beliefs intertwined managed to produce a greater range of knowledge in the human brain . With its experience a human can easily distinguish them said the same but have different meanings , or otherwise different words but the same meaning . Of all the advantages it has, on one hand, humans also have various limitations in managing large amounts of data , the limitations of doing a quick search , to the limitations in considering complex data . This is the basis of this study . The research in this paper will focus on an ambiguous word search prototype system , where the system is expected later to understand user queries performed more easily . SPARQL , RDF and ontology is the technology that will be used to support the course of this study .