Search

Heavy metals have been detected in coastal regions, in the vicinity of former mining sites. The level of these heavy metals frequently exceed the safe limits in what represents an important aquatic source of food for human consumption. A thorough and conventional study into the levels of contamination in the suspected area is a costly but necessary exercise in order to inform any appropriate mitigating actions. This study demonstrates the benefits of integrating scientific investigating methods comprising chemical analysis using aqua regia and geo-electrical surveys, i.e., electrical resistivity imaging and induced polarization, to attain the approximate levels of contamination for both lateral and vertical landscapes in an efficient way. The demonstration to present the conformity and reliability of this joint methods is at Saphan Hin, situated on the southeastern coast of Phuket, Thailand. The site is located downstream of Bang-Yai canal; a potential intermediary carrying heavy metals especially Lead and Zinc from a number of mines where the nearest is approximately 4 km away. Saphan Hin is an important place marine fishery to supply to an increasing number of tourists to Phuket. The combined results from using all methods in this study are consistent in showing that the extent of the over-the-limit contamination can be found up to at least 200 m from the estuary, and with the decreasing depths from around 20 m to 50 cm at further distances from mainland, demanding urgent measures for the health of local people as well as tourists. A model to adapt these monitoring methods for implementation in operating mines is proposed as an early warning system to prevent any leakage probably occurring under normal conditions.

Currently, the growth of information and technology is rapid. It makes a lot of things in various fields including education becoming more effective and efficient. In education, one of its implementation which is a game is proven to be a useful tool to support conventional teaching methods and bring a more natural understanding of learning materials. Moreover, as a product of the popular culture in modern society, video game mirrors the general culture practice in real life and reflects it via its own culture inside the game. This makes a video game can give a contribution to the social construction of reality as it affects the player’s view towards learning in real life and vice versa. From there, we see that there is an opportunity for learning about religion to be supported by its utilization to provide an interactive and fun learning experience. In this paper, we discuss how a video game is implemented to support religious learning. The game was developed with the scrum method where we surveyed to gather the user requirements before the design step. The game design was made by using the use case diagram and storyboard, and it was built using Unity version 2017.3.0f3.

Deep learning is applied to many research topics; Natural Language Processing, Image Processing, and Acoustic Recognition. In deep learning, neural networks have a very complex and deep structure and it is difficult to discuss why they work well or not. So you have to take a trial-and-error to improve their performances. We develop a mechanism to show how neural networks predict final results and help you to design a new neural network architecture based on its prediction criteria. Speaking concrete, we visualize important features to predict the final results with an attentional mechanism. In this paper, we take up sentient analysis, which is one of natural language processing tasks. In image processing visualizing weights of a neural network is a major approach and you can obtain intuitive results; object outlines and object components. However, in natural language processing, the approach is not interpretable because a discriminate function constructed by a neural network is a complex and nonlinear one and it is very difficult to correlate weights and words in a text. We employ Gated Convolutional Neural Network (GCNN) and introduce a self-attention mechanism to understand how GCNN determines sentiment polarities from raw reviews. GCNN can simulate an n-gram model and the self-attention mechanism can make correspondence between weights of a neural network and words clear. In experiments, we used Amazon reviews and evaluated the performance of the proposed method. Especially, the proposed method was able to emphasize some words in the review to determine sentiment polarity. Moreover, when the prediction was wrong, we were able to understand why the proposed method made mistakes because we found what words the proposed method emphasized.

Many locations in the world have been changing their main economic activity from agriculture to tourism pursuing more favorable opportunities for their development. Such a change in economic structure also leads to a significant alteration in the provincial greenhouse gas emissions. In order to investigate these differences, this study compares the greenhouse gases emitted from the major human activities in agriculture and tourism-based provinces. For this purpose, Phuket and Phayao provinces in Thailand were chosen as representatives for tourism and agriculture, respectively. Environmental and economic performances of other tourist and agricultural areas in different countries were also presented to examine commonalities and differences. It was found that in the agricultural provinces, the major part of greenhouse gas emissions came from rice cultivation, while in the tourist provinces, from the energy sector. The comparative results among several locations show that the tourist areas (especially urban-tourism-based areas) mostly had a higher eco-efficiency due to the large income from tourism and other services. They also had higher greenhouse gas indices than agricultural areas, due to energy consumption. The performance of agriculture-and-tourism-based provinces was outstanding with the lowest average carbon footprint per capita while not performing badly on the other indices, partly due to a decrease in biomass burning. At a time when global warming is the major world problem, agri-tourism can be considered to be an optimal choice and direction for the sustainable development of both agricultural and tourist provinces.

User's review on products and services is valuable information for both users and providers. The present paper conducted a polarity estimation of 73,589 reviews on hotels in Europe. Users rated one to five points for seven aspects (Value, Rooms, Location, Cleanliness, Check-in, Service, Business, Overall). In this paper, we predicted the polarity (positive/negative) of each aspect by using a machine learning method, SVM (Support Vector Machine), and feature selection, with more than 4 points being positive and less than 3 being negative. As a result, positive reviews with respect to six aspects, other than Business, were able to achieve 74% prediction performance (F-measure) with only 20 feature words. On the other hand, for negative reviews, optimal prediction performance could not be obtained unless almost all words were used, and on average F-measure was only 27%. The results indicate that positive reviews are simple, meanwhile negative reviews are diverse and hard to predict mechanically

Recurrent neural networks(RNNs) have been applied to sentiment classification but RNNs is usually heavier than convolutional neural networks (CNNs), turning more interest in the application of CNNs to language tasks. In this paper we propose a method to apply gated CNN (gCNN) with Maxpooling to sentiment classification of customer reviews. In our proposal, the application of gCNN is to sentiment classification, instead of constructing a language model. Our experiment is conducted with Amazon Product Review dataset and Japanese review dataset of TripAdvisor. The whole of each review is used as an input, instead of each sentence. The result is that a simple application of gCNN to sentiment classification achieved sufficient accuracies with the two datasets. Thus, an implication is that gCNN is proven to work fine for sentiment classification much faster than RNNs with fine results in the different language datasets.

Indonesia as the most populated country in ASEAN has inadequate health care treatment among ASEAN countries and building e-health for Indonesia is an urgent matter to strength Indonesian healthcare treatment. Moreover, there is a critical to strength national health information system which have challenges such as weak coordination among central, provincial and district health office and inadequate using of information system for decision making, where need to integrate all healthcare data such as health facility, health promotion, health financing, Human Resource for Health (HRH), community participation and health management. The development of Intelligent e-health will concern in healthcare and will build collaboration for data collection with public and private community health center and hospitals, Healthcare Social Security organizing body (Badan Penyelenggara Jaminan Sosial (BPJS) Kesehatan), the ministry of health of the Republic of Indonesia and World Health Organization (WHO). The development will implement current technology such as Data Warehouse, Big Data, Data mining, machine learning, knowledge management, Decision Support System and so on which dealing with structured and unstructured data. The development of Intelligent e-health will be built in smartphone application, using the Internet of Things (IoT) and including mining Indonesian healthcare information from social media.

Recurrent neural networks(RNNs) have been applied to sentiment classification but RNNs is usually heavier than convolutional neural networks (CNNs), turning more interest in the application of CNNs to language tasks. In this paper we propose a new method to apply gated CNN (gCNN) with Maxpooling via Spatial Pyramid Pooling (SPP) to sentiment classification of customer reviews. Our proposal consists of two parts. One is the application of gCNN to sentiment classification, instead of constructing a language model. The other is to incorporate SPP as pooling instead of a simple Maxpooling, which is originally proposed for image processing. Two experiments were conducted to evaluate our proposal. The result shows that gCNN works better than Support Vector Machines, which suggests that advanced CNNs can be effectively employed for various language processing tasks. On the other hand, SPP did not work well for our taks and the simple Maxpooling to the whole convolutional output performed the best. This suggests that Maxpooling works fine also with document classification tasks while the simple application of SPP for document classification is not so appropriate.

Word segmentation is one of the most fundamental processes for most natural language processing tasks. In particular, languages with no word boundary in writing such as Chinese, Japanese, Korean, Thai, and Burmese need it. However, the Burmese language still waits for a technique with good performance. In this paper, we propose a new technique for Burmese word segmentation employing the idea of Character Clustering for Conditional Random Fields. Character clusters are groups of some inseparable characters due to language characteristics. We proposed a set of 29 types of Burmese Character Clusters (BCCs) as rules, and Conditional Random Fields is applied as a sequential labelling machine learning method. We compared our proposed method with CRF without BCC and Syllable-based CRFs. The result shows that our proposed method achieved the highest performance.

Generally, evaluations in customer reviews are shown with text and score. In addition, there are 2 kinds of scored evaluations, which are a comprehensive evaluation and topic evaluations. However, topic evaluations are not often given by reviewers. Therefore, it is required to estimate the evaluations. In this paper, we researched a performance of estimation by SVM. Furthermore, we compared results of estimation by SVM and scanning with dependency relation tree. The result shew that precision of estimation by each method is low. On the other hand, the result shew that precision in positive score is over about 70% in each topic.

Recently, the customer reviews became many important and useful information about evaluation and/or needs of customers. Many research about sentiment analysis for the customer reviews were proposed aggressively. In this paper, we focus on evaluation sentence patterns and the words appearing in them to estimate sentiment values of texts, and compared with a Bag-of-Words-based SVM classification. We investigated how many feature words obtained by SVM are included in the words extracted from the sentences of evaluation sentence patterns, and roughly 75% are included, which suggests that to just focus on a limited number sentences of evaluation sentence patterns can produce a comparable result even in its early stage.