• Journal of Intelligence and Information Systems
• /
• v.16 no.4
• /
• pp.195-211
• /
• 2010

An exhibition is defined as market events for specific duration to present exhibitors' main product range to either business or private visitors, and it also plays a key role as effective marketing channels. Especially, as the effect of the opinions of the visitors after the exhibition impacts directly on sales or the image of companies, exhibition organizers must consider various needs of visitors. To meet needs of visitors, ubiquitous technologies have been applied in some exhibitions. However, despite of the development of the ubiquitous technologies, their services cannot always reflect visitors' preferences as they only generate information when visitors request. As a result, they have reached their limit to meet needs of visitors, which consequently might lead them to loss of marketing opportunity. Recommendation systems can be the right type to overcome these limitations. They can recommend the booths to coincide with visitors' preferences, so that they help visitors who are in difficulty for choices in exhibition environment. One of the most successful and widely used technologies for building recommender systems is called Collaborative Filtering. Traditional recommender systems, however, only use neighbors' evaluations or behaviors for a personalized prediction. Therefore, they can not reflect visitors' dynamic preference, and also lack of accuracy in exhibition environment. Although there is much useful information to infer visitors' preference in ubiquitous environment (e.g., visitors' current location, booth visit path, and so on), they use only limited information for recommendation. In this study, we propose a booth recommendation methodology using Sequential Association Rule which considers the sequence of visiting. Recent studies of Sequential Association Rule use the constraints to improve the performance. However, since traditional Sequential Association Rule considers the whole rules to recommendation, they have a scalability problem when they are adapted to a large exhibition scale. To solve this problem, our methodology composes the confidence database before recommendation process. To compose the confidence database, we first search preceding rules which have the frequency above threshold. Next, we compute the confidences of each preceding rules to each booth which is not contained in preceding rules. Therefore, the confidence database has two kinds of information which are preceding rules and their confidence to each booth. In recommendation process, we just generate preceding rules of the target visitors based on the records of the visits, and recommend booths according to the confidence database. Throughout these steps, we expect reduction of time spent on recommendation process. To evaluate proposed methodology, we use real booth visit records which are collected by RFID technology in IT exhibition. Booth visit records also contain the visit sequence of each visitor. We compare the performance of proposed methodology with traditional Collaborative Filtering system. As a result, our proposed methodology generally shows higher performance than traditional Collaborative Filtering. We can also see some features of it in experimental results. First, it shows the highest performance at one booth recommendation. It detects preceding rules with some portions of visitors. Therefore, if there is a visitor who moved with very a different pattern compared to the whole visitors, it cannot give a correct recommendation for him/her even though we increase the number of recommendation. Trained by the whole visitors, it cannot correctly give recommendation to visitors who have a unique path. Second, the performance of general recommendation systems increase as time expands. However, our methodology shows higher performance with limited information like one or two time periods. Therefore, not only can it recommend even if there is not much information of the target visitors' booth visit records, but also it uses only small amount of information in recommendation process. We expect that it can give real?time recommendations in exhibition environment. Overall, our methodology shows higher performance ability than traditional Collaborative Filtering systems, we expect it could be applied in booth recommendation system to satisfy visitors in exhibition environment.

• Korean journal of applied entomology
• /
• v.24 no.3 s.64
• /
• pp.117-121
• /
• 1985

Through the laboratory and vinyl house experiments, the effects of inoculum density, plant age and temperature on the incidence of Phytophthora crown rot of pepper (Capsicum annum L.) were investigated. The propagule survival was greater in the natural soil than in autoclaved soil within first 2 weeks when the sporangial suspension of the pathogenic fungus was incorporated into soil, thereafter the survivability reduced rapidly. The propagule was not detectable in 35 days by means of Papavizas selective medium neither in natural nor in autoclaved soil. At least 5 sporangia per gram soil were required to induce crown rot for 30 days old pepper seedlings. Further increase in inoculum concentration above this threshold level resulted in higher disease incidence and shorter incubation period. When the same amount of inoculum was infested, higher disease incidence was observed for younger plants until 3 weeks after inoculation. On the other hand after 4 weeks this tendency was not extended any more. Younger plants were recognized as having shorter incubation period upon infection, however, the days from first symptom appearance to complete death were not significantly different among differently aged seedlings. Exposure of inoculated pepper seedlings to $25^{\circ}C$ resulted in highest infection rates and followed by those to $30^{\circ}C\;and\;20^{\circ}C$ but no disease was found at $15^{\circ}C\;and\;35^{\circ}C$ for 10 days. When the plants previously incubated at different temperature for 10 days were moved to $25^{\circ}C$ room temperature, prior exposure to $20^{\circ}C\;and\;30^{\circ}C$ brought continuous disease development. Even those plants preincubated at $15^{\circ}C$ were diseased up to 50%. But the prior exposure to $35^{\circ}C$ induced no symptom developed, indicating no seedlings infected at all.

• Korean journal of applied entomology
• /
• v.49 no.4
• /
• pp.305-312
• /
• 2010

The developmental time of immature stages of Paromius exiguus (Distant) was investigated at nine constant temperatures (15, 17.5, 20, 22.5, 25, 27.5, 30, 32.5, $35{\pm}1^{\circ}C$), 20-30% RH, and a photoperiod of 14:10h (L:D). Eggs did not develop at $15^{\circ}C$, and their developmental time decreased with increasing temperatures. Its developmental time was longest at $17.5^{\circ}C$ (28.2 days) and shortest at $35^{\circ}C$ (5.9 days). The first nymphs failed to reach the next nymphal stage at 17.5 and $35^{\circ}C$. Nymphal developmental time decreased with increasing temperatures between $20^{\circ}C$ and $32.5^{\circ}C$, and developmental rate was decreased at temperatures above $30^{\circ}C$ in all stages except for the fourth nymphal stage. The relationship between developmental rate and temperature fit a linear model and three nonlinear models (Briere 1, Lactin 2, and Logan 6). The lower threshold temperature of egg and total nymphal stage was $l3.8^{\circ}C$ and $15.3^{\circ}C$, respectively. The thermal constant required to reach complete egg and the total nymphal stage was 109.9 and 312.5DD, respectively. The Logan-6 model was best fitted ($r^2$=0.94-0.99), among three nonlinear models. The distribution of completion of each development stage was well described by the 3-parameter Weibull function ($r^2$=0.91-0.99).

• Journal of Oral Medicine and Pain
• /
• v.30 no.1
• /
• pp.87-106
• /
• 2005

In order to evaluate the mechanism of transmission as well as processing of sensory information originating from low-threshold mechanoreceptor in oral and maxillofacial region at primary synaptic region of trigeminal nervous system, vibrissa afferent fibers of adult cat were labeled with intra-axonal HRP injection. Serial sections containing labeled boutons were obtained from the piece of trigeminal interpolar nucleus. Under electron microscope, total 30 labeled boutons were observed, and ultrastructural characteristics, frequency of occurence, synaptic organizations of vibrissa afferent terminals were analysed. The results were as follows: 1. Labeled boutons contained clear, spherical synaptic vesicles with diameter of 45$\sim$55nm. They formed asymmetrical synapse with dendrites showing definite postsynaptic density, larger synaptic cleft, multiple synaptic structures at various regions. With unlabeled axon terminals(p-ending) containing polymorphic synaptic vesicles, they formed symmetrical synapse showing indefinite postsynaptic density and narrower synaptic area. 2. Each labeled bouton formed 1 to 15 synapses, the average of 4.77$\pm$3.37 contacts per labeled bouton, with adjacent neuronal profiles. Relatively complex synaptic organization, which formed synapses with more than 5 neuronal profiles, was observed in a large number(46.7%, n=14) of labeled boutons. 3. Axo-somatic synapse was not observed. The number of axo-dendritic synapse was 1.83$\pm$1.37 per labeled bouton. Majority(85.0%) of axo-dendritic synapses were formed with dendritic shafts, nonprimary dendrites(n=47, 1.57$\pm$1.38/1 bouton), however, synapses formed with primary dendrites(n=6, 0.20$\pm$0.41/1 bouton) or dendritic spines(n=2, 0.07$\pm$0.25/1 bouton) were rare. 4. 76.7%(n=23) of labeled boutons formed axo-axonic synapse (2.93$\pm$2.36/1 bouton) with p-endings containing pleomorphic vesicles. Synaptic triad, in which p-endings formed synapses with labeled boutons and dendrites adjacent to the labeled boutons simultaneoulsy, were also observed in 60.0%(n=18) of labeled boutons. From the above results, vibrissa afferent terminals of adult cat showed distinctive synaptic organization in the trigeminal interpolar nucleus, thus, suggests their correlation with the function of the trigeminal interpolaris nucleus, which participates in processing of complex sensory information such as two-point discrimination and motivational-affective action. Further studies on physiologic functions such as quantitative analysis on ultrastructures of afferent terminals and nerve transmitters participating in presynaptic inhibition are required.

• Asia pacific journal of information systems
• /
• v.23 no.2
• /
• pp.1-20
• /
• 2013

Social commerce has been gaining explosive popularity, with typical examples of the model such as Groupon and Level Up. Both local business owners and consumers can benefit from this new e-commerce model. Local business owners have a chance to access potential customers and promote their products in a way that could not have otherwise been easily possible, and consumers can enjoy discounted offerings. However, questions have been increasingly raised about the value and future of the social commerce model. A recent survey shows that about a third of 324 business owners who ran a daily-deal promotion in Groupon went behind. Furthermore, more than half of the surveyed merchants did not express enthusiasm about running the promotion again. The same goes for the case in Korea, where more than half of the surveyed clients reported no significant change or even decrease in profits compared to before the use of social commerce model. Why do local business owners fail to exploit the benefits from the promotions and advertisements through the social commerce model and to make profits? Without answering this question, the model would fall under suspicion and even its sustainability might be challenged. This study aims to look into problems in the current social commerce transactions and provide implications for the social commerce model, so that the model would get a foothold for next growth. Drawing on justice theory, this study develops theoretical arguments for the effects of transaction characteristics on consumers' distributive justice and purchase intention in the social commerce. Specifically, this study focuses on two characteristics of social commerce transactions-the discount rate and the purchase rate of products-and investigates their effects on consumers' perception of distributive justice for discounted transactions in the social commerce and their perception of distributive justice for regular-priced transactions. This study also examines the relationship between distributive justice and purchase intention. We conducted an online experiment and gathered data from 115 participants to test the hypotheses. Each participant was randomly assigned to one of nine manipulated scenarios of social commerce transactions, which were generated based on the combination of three levels of purchase rate (high, medium, and low) and three levels of discount rate (high, medium, and low). We conducted MANOVA and post-hoc ANOVA to test hypotheses about the relationships between the transaction characteristics (purchase rate and discount rate) and distributive justice for each of the discounted transaction and the regular-priced transaction. We also employed a PLS analysis to test relations between distributive justice and purchase intentions. Analysis results show that a higher discount rate increases distributive justice for the discounted transaction but decreases distributive justice for the regular-priced transaction. This, coupled with the result that distributive justice for each type of transaction has a positive effect on the corresponding purchase intention, implies that a large discount in the social commerce may be helpful for attracting consumers, but harmful to the business after the promotion. However, further examination reveals curvilinear effects of the discount rate on both types of distributive justice. Specifically, we find distributive justice for the discounted transaction increases concavely as the discount rate increases while distributive justice for the regular-priced transaction decreases concavely with the dscount rate. This implies that there exists an appropriate discount rate which could promote the discounted transaction while not hurting future business of regular-priced transactions. Next, the purchase rate is found to be a critical factor that facilitates the regular-priced transaction. It has a convexly positive influence on distributive justice for the transaction. Therefore, an increase of the rate beyond some threshold would lead to a substantial level of distributive justice for the regular-priced transaction, threrby boosting future transactions. This implies that social commerce firms and sellers should employ various non-price stimuli to promote the purchase rate. Finally, we find no significant relationship between the purchase rate and distributive justice for the discounted transaction. Based on the above results, we provide several implications with future research directions.

• Journal of Korean Society of Forest Science
• /
• v.110 no.4
• /
• pp.504-515
• /
• 2021

To examine the relationship between environmental factors influencing the habitat of Aconitum austrokoreense Koidz., this study employed the MexEnt model to evaluate 21 environmental factors. Fourteen environmental factors having an AUC of at least 0.6 were found to be the age of stand, growing stock, altitude, topography, topographic wetness index, solar radiation, soil texture, mean temperature in January, mean temperature in April, mean annual temperature, mean rainfall in January, mean rainfall in August, and mean annual rainfall. Based on the response curves of the 14 descriptive factors, Aconitum austrokoreense Koidz. on the Baekun Mountain were deemed more suitable for sites at an altitude of 600 m or lower, and habitats were not significantly affected by the inclination angle. The preferred conditions were high stand density, sites close to valleys, and distribution in the northwestern direction. Under the five-age class system, the species were more likely to be observed for lower classes. The preferred solar radiation in this study was 1.2 MJ/m². The species were less likely to be observed when the topographic wetness index fell below the reference value of 4.5, and were more likely observed above 7.5 (reference of threshold). Soil analysis showed that Aconitum austrokoreense Koidz. was more likely to thrive in sandy loam than clay. Suitable conditions were a mean January temperature of - 4.4℃ to -2.5℃, mean April temperature of 8.8℃-10.0℃, and mean annual temperature of 9.6℃-11.0℃. Aconitum austrokoreense Koidz. was first observed in sites with a mean annual rainfall of 1,670- 1,720 mm, and a mean August rainfall of at least 350 mm. Therefore, sites with increasing rainfall of up to 390 mm were preferred. The area of potential habitats having distributive significance of 75% or higher was 202 ha, or 1.8% of the area covered in this study.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.16 no.2
• /
• pp.51-78
• /
• 2021

Followed by 'start-up', the theme of 'scale-up' has been considered as an important agenda in both corporate and policy spheres. In particular, although it is a term commonly used in industry and policy fields, even a conceptual definition has not been achieved from the academic perspective. "Corporate Growth" in the academic aspect and "Business Growth" in the practical management field have different understandings (Achtenhagen et al., 2010). Previous research on corporate growth has not departed from Penrose(1959)'s "Firm as a bundle of resources" and "the role of managers". Based on the theory and background of economics, existing research has mainly examined factors that contribute to firms' growth and their growth patterns. Comparatively, we lack knowledge on the firms' growth with a focus on 'annual revenue growth rate'. In the early stage of the firms, they tend to exhibit a high growth rate as it started with a lower level of annual revenue. However, when the firms reach annual revenue of more than 100 billion KRW, a threshold to be classified as a 'middle-standing enterprise' by Korean standards, they are unlikely to reach a high level of revenue growth rate. In our study, we used our sample of 333 companies (6.7% out of 5,000 'fastest-growing' companies) which reached 15% of the compound annual growth rate in the last three years with more than USD 100 million. It shows that sustaining 'high-growth' above a certain firm size is difficult. The study focuses on firms with annual revenue of more than $100 billion (approximately 120 billion KRW) from the 'Inc. 2020 fast-growing companies 5,000' list. The companies have been categorized into 1) Fast-growing companies (revenue CAGR 15%~40% between 2016 and 2019), 2) Hyper-growing companies (40%~99.9%), and 3) Super-growing (100% or more) with in-depth analysis of each group's characteristics. Also, the relationship between the revenue growth rate, individual company's strategy choice (market orientation, generic strategy, growth strategy, pioneer strategy), industry/market environment, and firm age is investigated with a quantitative approach. Through conducting the study, it aims to provide a reference to the 'Hyper-Growing Model' that combines the paths and factors of growth strategies. For policymakers, our study intends to provide a reference to which factors or environmental variables should be considered for 'optimal effective combinations' to promote firms' growth.

• Journal of Nutrition and Health
• /
• v.57 no.3
• /
• pp.292-306
• /
• 2024

Purpose: This study investigated fruit and vegetable intake and associated socio-demographic and dietary factors, and compared the nutritional intake according to the fruit and vegetable intake level among Korean adolescents. Methods: This study was conducted on 1,676 adolescents who participated in the 2016-2019 Korea National Health and Nutrition Examination Survey. The subjects were classified into four groups based on the fruit and vegetable intake recommendations in 2020 Dietary Reference Intakes for Koreans: Application (KDRIs Application): sufficient fruit intake (SF) group, sufficient vegetables intake (SV) group, sufficient fruit and vegetables intake (SFV) group, and not sufficient fruit and vegetable intake (NS) group The nutrient intake per day in each group was compared.. Logistic regression analysis was performed to examine the factors influencing fruit and vegetables intake. Results: In the sample of adolescents surveyed, only 1.40% met the recommended daily intake of fruits and vegetables, while 79.54% fell below the established threshold for adequate consumption. Female adolescents, those with fathers holding university degrees or above, and those who ate breakfast at least three times a week were likelier to have adequate fruit intake. Male adolescents and those from higher-income households were likelier to consume vegetables. Females, those who ate out daily, those from lower-income households, and those who understood food labels were likelier to have adequate fruit and vegetable intake. The daily nutrient intake and intake-to-requirement ratio significantly differed according to the fruit and vegetable intake groups. The NS and SF group had lower ratios for calcium and iron, while the NS group had the lowest vitamin A and C intake. By contrast, the SFV group met almost all daily nutrient requirements, except for calcium and vitamin A. Conclusion: This study highlights the need for nutrition education programs to encourage adolescents to consume adequate amounts of fruits and vegetables.

• The Journal of Korean Academy of Prosthodontics
• /
• v.40 no.1
• /
• pp.1-17
• /
• 2002

The purpose of this study is to examine the possibility of thermal injury to bone tissues during an implant site preparation under the same condition as a typical clinical practice of $Br{\aa}nemark$ implant system. All the burs for $Br{\aa}nemark$ implant system were studied except the round bur The experiments involved 880 drilling cases : 50 cases for each of the 5 steps of NP, 5 steps of RP, and 7 steps of WP, all including srew tap, and 30 cases of 2mm twist drill. For precision drilling, a precision handpiece restraining system was developed (Eungyong Machinery Co., Korea). The system kept the drill parallel to the drilling path and allowed horizontal adjustment of the drill with as little as $1{\mu}m$ increment. The thermocouple insertion hole. that is 0.9mm in diameter and 8mm in depth, was prepared 0.2mm away from the tapping bur the last drilling step. The temperatures due to countersink, pilot drill, and other drills were measured at the surface of the bone, at the depths of 4mm and 8mm respectively. Countersink drilling temperature was measured by attaching the tip of a thermocouple at the rim of the countersink. To assure temperature measurement at the desired depths, 'bent-thermocouples' with their tips of 4 and 8mm bent at $120^{\circ}$ were used. The profiles of temperature variation were recorded continuously at one second interval using a thermometer with memory function (Fluke Co. U.S.A.) and 0.7mm thermocouples (Omega Co., U.S.A.). To simulate typical clinical conditions, 35mm square samples of bovine scapular bone were utilized. The samples were approximately 20mm thick with the cortical thickness on the drilling side ranging from 1 to 2mm. A sample was placed in a container of saline solution so that its lower half is submerged into the solution and the upper half exposed to the room air, which averaged $24.9^{\circ}C$. The temperature of the saline solution was maintained at $36.5^{\circ}C$ using an electric heater (J. O Tech Co., Korea). This experimental condition was similar to that of a patient s opened mouth. The study revealed that a 2mm twist drill required greatest attention. As a guide drill, a twist drill is required to bore through a 'virgin bone,' rather than merely enlarging an already drilled hole as is the case with other drills. This typically generates greater amount of heat. Furthermore, one tends to apply a greater pressure to overcome drilling difficulty, thus producing even greater amount heat. 150 experiments were conducted for 2mm twist drill. For 140 cases, drill pressure of 750g was sufficient, and 10 cases required additional 500 or 100g of drilling pressure. In case of the former. 3 of the 140 cases produced the temperature greater than $47^{\circ}C$, the threshold temperature of degeneration of bone tissue (1983. Eriksson et al.) which is also the reference temperature in this study. In each of the 10 cases requiring extra pressure, the temperature exceeded the reference temperature. More significantly, a surge of heat was observed in each of these cases This observations led to addtional 20 drilling experiments on dense bones. For 10 of these cases, the pressure of 1,250g was applied. For the other 10, 1.750g were applied. In each of these cases, it was also observed that the temperature rose abruptly far above the thresh old temperature of $47^{\circ}C$, sometimes even to 70 or $80^{\circ}C$. It was also observed that the increased drilling pressure influenced the shortening of drilling time more than the rise of drilling temperature. This suggests the desirability of clinically reconsidering application of extra pressures to prevent possible injury to bone tissues. An analysis of these two extra pressure groups of 1,250g and 1,750g revealed that the t-statistics for reduced amount of drilling time due to extra pressure and increased peak temperature due to the same were 10.80 and 2.08 respectively suggesting that drilling time was more influenced than temperature. All the subsequent drillings after the drilling with a 2mm twist drill did not produce excessive heat, i.e. the heat generation is at the same or below the body temperature level. Some of screw tap, pilot, and countersink showed negative correlation coefficients between the generated heat and the drilling time. indicating the more the drilling time, the lower the temperature. The study also revealed that the drilling time was increased as a function of frequency of the use of the drill. Under the drilling pressure of 750g, it was revealed that the drilling time for an old twist drill that has already drilled 40 times was 4.5 times longer than a new drill The measurement was taken for the first 10 drillings of a new drill and 10 drillings of an old drill that has already been used for 40 drillings. 'Test Statistics' of small samples t-test was 3.49, confirming that the used twist drills require longer drilling time than new ones. On the other hand, it was revealed that there was no significant difference in drilling temperature between the new drill and the old twist drill. Finally, the following conclusions were reached from this study : 1 Used drilling bur causes almost no change in drilling temperature but increase in drilling time through 50 drillings under the manufacturer-recommended cooling conditions and the drilling pressure of 750g. 2. The heat that is generated through drilling mattered only in the case of 2mm twist drills, the first drill to be used in bone drilling process for all the other drills there is no significant problem. 3. If the drilling pressure is increased when a 2mm twist drill reaches a dense bone, the temperature rises abruptly even under the manufacturer-recommended cooling conditions. 4. Drilling heat was the highest at the final moment of the drilling process.

• Journal of Intelligence and Information Systems
• /
• v.19 no.3
• /
• pp.113-125
• /
• 2013

Every company wants to know customer's requirement and makes an effort to meet them. Cause that, communication between customer and company became core competition of business and that important is increasing continuously. There are several strategies to find customer's needs, but VOC (Voice of customer) is one of most powerful communication tools and VOC gathering by several channels as telephone, post, e-mail, website and so on is so meaningful. So, almost company is gathering VOC and operating VOC system. VOC is important not only to business organization but also public organization such as government, education institute, and medical center that should drive up public service quality and customer satisfaction. Accordingly, they make a VOC gathering and analyzing System and then use for making a new product and service, and upgrade. In recent years, innovations in internet and ICT have made diverse channels such as SNS, mobile, website and call-center to collect VOC data. Although a lot of VOC data is collected through diverse channel, the proper utilization is still difficult. It is because the VOC data is made of very emotional contents by voice or text of informal style and the volume of the VOC data are so big. These unstructured big data make a difficult to store and analyze for use by human. So that, the organization need to automatic collecting, storing, classifying and analyzing system for unstructured big VOC data. This study propose an intelligent VOC analyzing system based on opinion mining to classify the unstructured VOC data automatically and determine the polarity as well as the type of VOC. And then, the basis of the VOC opinion analyzing system, called domain-oriented sentiment dictionary is created and corresponding stages are presented in detail. The experiment is conducted with 4,300 VOC data collected from a medical website to measure the effectiveness of the proposed system and utilized them to develop the sensitive data dictionary by determining the special sentiment vocabulary and their polarity value in a medical domain. Through the experiment, it comes out that positive terms such as "칭찬, 친절함, 감사, 무사히, 잘해, 감동, 미소" have high positive opinion value, and negative terms such as "퉁명, 뭡니까, 말하더군요, 무시하는" have strong negative opinion. These terms are in general use and the experiment result seems to be a high probability of opinion polarity. Furthermore, the accuracy of proposed VOC classification model has been compared and the highest classification accuracy of 77.8% is conformed at threshold with -0.50 of opinion classification of VOC. Through the proposed intelligent VOC analyzing system, the real time opinion classification and response priority of VOC can be predicted. Ultimately the positive effectiveness is expected to catch the customer complains at early stage and deal with it quickly with the lower number of staff to operate the VOC system. It can be made available human resource and time of customer service part. Above all, this study is new try to automatic analyzing the unstructured VOC data using opinion mining, and shows that the system could be used as variable to classify the positive or negative polarity of VOC opinion. It is expected to suggest practical framework of the VOC analysis to diverse use and the model can be used as real VOC analyzing system if it is implemented as system. Despite experiment results and expectation, this study has several limits. First of all, the sample data is only collected from a hospital web-site. It means that the sentimental dictionary made by sample data can be lean too much towards on that hospital and web-site. Therefore, next research has to take several channels such as call-center and SNS, and other domain like government, financial company, and education institute.