Teses / dissertações sobre o tema "Application Shiny"

Les interactions entre espèces dans les communautés écologiques sont complexes : de nombreuses espèces peuvent interagir les unes avec les autres de façons variées et à différentes échelles spatiales et temporelles. De plus, ces réseaux d'interactions sont la résultante de multiples causes, engendrent de multiples processus et ont des conséquences parfois indirectes transmises au travers de la structure du réseau. Cette complexité nécessite une diversité d'approches pour comprendre les déterminants des interactions et prédire leurs effets dans les systèmes écologiques. Cette thèse étudie plusieurs aspects des réseaux d'interactions écologiques par une approche méthodologique qui se concentre sur la description, l'évaluation et le développement de méthodes statistiques et d'outils logiciels. Dans une première partie, j'étudie les causes de la structure des réseaux d'interactions en me concentrant sur les niches d'interactions et en utilisant la notion de concordance des traits entre espèces. Pour cela, j'utilise des méthodes de la famille de l'analyse des correspondances et j'applique et j'étends des méthodes de mise à l'échelle réciproque à l'analyse de réseaux bipartites. J'applique ces méthodes à l'analyse d'un réseau d'interactions plantes-frugivores d'une forêt de montagne péruvienne et je montre que les traits des espèces peuvent être reliés à leur largeur de niche. Dans une deuxième partie, j'étudie les conséquences des interactions au travers de leur influence sur la répartition spatio-temporelle des espèces. Pour cela, j'utilise des processus de Hawkes multivariés pour analyser des données de pièges photographiques. J'illustre ces modèles sur cinq mammifères de la savane sud-africaine et je montre des attractions et évitements entre plusieurs de ces espèces à courte échelle spatio-temporelle. Dans une troisième partie, je me penche sur l'analyse de données collectées par pièges photographiques. Je développe un package R pour nettoyer et standardiser ces données à l'usage du programme Snapshot Safari, ainsi qu'une application Shiny destinée à un usage plus général pour visualiser de données de façon interactive et reproductible. Cette thèse présente des méthodes statistiques et outils logiciels pour analyser des données écologiques complexes et améliorer la compréhension des réseaux d'interactions. Ces résultats ouvrent des perspectives nouvelles concernant l'analyse de données écologiques ainsi que les développement méthodologique en écologie
Interactions between species in ecological communities are complex: many species can interact with each other in a variety of ways and at different spatial and temporal scales. Moreover, these interaction networks are the result of multiple causes, generate multiple processes and can have indirect effects transmitted through the structure of the network. This complexity calls for a variety of approaches to understand the determinants of interactions and predict their effects in ecological systems. This thesis studies several aspects of ecological interaction networks using a methodological approach that focuses on the description, evaluation and development of statistical methods and software tools. In a first part, I study causes of the structure of interaction networks, focusing on interaction niches and using the notion of trait matching between species. To this end, I use methods from the correspondence analysis family and apply and extend reciprocal scaling methods to the analysis of bipartite networks. I apply these methods to the analysis of a plant-frugivore interaction network in a Peruvian montane forest, and show that species traits can be related to their niche width. In a second part, I study the consequences of interactions through their influence on the spatio-temporal distribution of species. To this end, I use multivariate Hawkes processes to analyze camera trap data. I illustrate these models on five mammals from the South African savanna, showing attraction and avoidance between several of these species at a short spatio-temporal scale. In a third part, I consider camera trap data analysis. I develop a R package to clean and standardize camera trap data intended for the Snapshot Safari program, as well as a Shiny application intended for a more general use to visualize data in an interactive and reproducible way. This thesis presents statistical methods and software tools to analyze complex ecological data and improve our understanding of interaction networks. These results open new perspectives on ecological data analysis and methodological development in ecology

Plusieurs évolutions sont constatées dans la recherche en biologie. Tout d’abord, les études menées reposent souvent sur des approches expérimentales quantitatives. L’analyse et l’interprétation des résultats requièrent l’utilisation de l’informatique et des statistiques. Également, en complément des études centrées sur des objets biologiques isolés, les technologies expérimentales haut débit permettent l’étude des systèmes (caractérisation des composants du système ainsi que des interactions entre ces composants). De très grandes quantités de données sont disponibles dans les bases de données publiques, librement réutilisables pour de nouvelles problématiques. Enfin, les données utiles pour les recherches en biologie sont très hétérogènes (données numériques, de textes, images, séquences biologiques, etc.) et conservées sur des supports d’information également très hétérogènes (papiers ou numériques). Ainsi « l’analyse de données » s’est petit à petit imposée comme une problématique de recherche à part entière et en seulement une dizaine d’années, le domaine de la « Bioinformatique » s’est en conséquence totalement réinventé. Disposer d’une grande quantité de données pour répondre à un questionnement biologique n’est souvent pas le défi principal. La vraie difficulté est la capacité des chercheurs à convertir les données en information, puis en connaissance. Dans ce contexte, plusieurs problématiques de recherche en biologie ont été abordées lors de cette thèse. La première concerne l’étude de l’homéostasie du fer chez la levure pathogène Candida glabrata. La seconde concerne l’étude systématique des modifications post-traductionnelles des protéines chez la levure pathogène Candida albicans. Pour ces deux projets, des données « omiques » ont été exploitées : transcriptomiques et protéomiques. Des outils bioinformatiques et des outils d’analyses ont été implémentés en parallèle conduisant à l’émergence de nouvelles hypothèses de recherche en biologie. Une attention particulière et constante a aussi été portée sur les problématiques de reproductibilité et de partage des résultats avec la communauté scientifique
Biological research is changing. First, studies are often based on quantitative experimental approaches. The analysis and the interpretation of the obtained results thus need computer science and statistics. Also, together with studies focused on isolated biological objects, high throughput experimental technologies allow to capture the functioning of biological systems (identification of components as well as the interactions between them). Very large amounts of data are also available in public databases, freely reusable to solve new open questions. Finally, the data in biological research are heterogeneous (digital data, texts, images, biological sequences, etc.) and stored on multiple supports (paper or digital). Thus, "data analysis" has gradually emerged as a key research issue, and in only ten years, the field of "Bioinformatics" has been significantly changed. Having a large amount of data to answer a biological question is often not the main challenge. The real challenge is the ability of researchers to convert the data into information and then into knowledge. In this context, several biological research projects were addressed in this thesis. The first concerns the study of iron homeostasis in the pathogenic yeast Candida glabrata. The second concerns the systematic investigation of post-translational modifications of proteins in the pathogenic yeast Candida albicans. In these two projects, omics data were used: transcriptomics and proteomics. Appropriate bioinformatics and analysis tools were developed, leading to the emergence of new research hypotheses. Particular and constant attention has also been paid to the question of data reproducibility and sharing of results with the scientific community

碩士
國立高雄大學
統計學研究所
107
This study uses the Shiny package in the R programming language to create an interactive web application for data science including time series, data mining, machine learning, regression analysis, etc. Since the Shiny package is a dynamic and interactive application that allows users to change settings and check results right away, it saves users a lot of time from rerunning and modifying programs. In addition, this interactive web application is very helpful for the introduction and further understanding of statistical methods. Users can describe the data through interactive methods and use R programming language to complete all operations, which is helpful for improving the effectiveness of data science in teaching and learning.

Tese de mestrado, Ciências Biofarmacêuticas, Universidade de Lisboa, Faculdade de Farmácia, 2018
A farmacoeconomia é uma disciplina que avalia o uso de medicamentos em termos de recursos na maximização da saúde da população. Dado que os recursos para os cuidados de saúde são finitos, a avaliação económica envolve a estimativa do custo de oportunidade, i.e., os benefícios marginais perdidos como resultado do deslocamento de tratamentos ou serviços existentes para financiar novos medicamentos. A farmacocinética é a ciência que visa o estudo do movimento de fármacos no organismo, o que inclui a absorção, distribuição, metabolismo e eliminação destes e seus metabolitos. Com o advento da química analítica e métodos de quantificação sofisticados, bem como de um aumento do poder de computação, a farmacocinética como ciência tem tido um desenvolvimento exponencial. Uma das áreas da farmacocinética que se tem desenvolvido mais é a farmacocinética populacional: apesar da farmacocinética de um fármaco poder ser estudada individualmente em cada indivíduo, a abordagem populacional é benéfica para o estudo de grupos de pacientes que são difíceis de investigar, como a população de bebés prematuros, pacientes com insuficiência hepática ou renal. Na farmacocinética populacional, cada indivíduo é avaliado simultaneamente com o modelo de efeitos mistos não-lineares (parametrização). Não linear significa que a variável dependente dessa concentração está relacionada não linearmente à associação de variáveis independentes e parâmetros do modelo. Efeitos fixos refere-se aos parâmetros que não se alteram em indivíduos, enquanto o efeito aleatório se refere àqueles parâmetros que se alteram através dos indivíduos. O principal objetivo das estimativas de modelação farmacocinética populacional é o de procurar os parâmetros de farmacocinética populacional e fonte de variabilidade. Os objetivos restantes consistem em concentrações observadas da dose administrada pela deteção das covariáveis preditivas na população avaliada. Em farmacocinética populacional, os indivíduos poderão apenas fornecer dados de concentração plasmática escassos. As cinco principais partes fundamentais para a construção de um modelo farmacocinético populacional incluem: dados, modelo estrutural, modelo estatístico, modelo de covariáveis e software de modelação. Os modelos estruturais definem o perfil de concentração plasmática ao longo do tempo nos indivíduos. Os modelos estatísticos descrevem a variabilidade aleatória na população que não é explicável (como a variabilidade entre as ocasiões), entre a variabilidade do indivíduo ou a variabilidade residual. Os modelos de covariável demonstram a variabilidade estimada pelas características da população, como covariáveis. O software de modelação, como o software de modelação de efeitos mistos não linear, permite a combinação de dados e modelos e aplica o método de estimativa para avaliar parâmetros para os modelos estatísticos, estruturais e de covariáveis que definem os dados. Na modelação farmacocinética populacional, o software possui um algoritmo de minimização do valor da função objetivo, praticando a estimativa de máxima verossimilhança. No momento da adaptação dos dados populacionais, a concentração estimada para cada indivíduo é influenciada pela variância dos parâmetros populacionais e de cada parâmetro individual, e pela variação em cada valor das concentrações previstas e observadas. A avaliação da probabilidade marginal depende dos parâmetros de efeito aleatório (η) e efeito fixo da população. Não há existência de solução analítica para verossimilhança marginal. Enquanto buscava a máxima verossimilhança, inúmeras abordagens foram aplicadas para a aproximação da verossimilhança marginal. O FOCE e o LAPLACE são as abordagens mais antigas que estimam a verdadeira verossimilhança com uma função adicional simplificada. O trabalho de dissertação no âmbito do Mestrado em Ciências Biofarmacêuticas teve por objetivo o estabelecimento de ferramentas baseadas em simulação de dados com base em modelos farmacocinéticos populacionais para uma posterior análise farmacoeconómica. Neste trabalho utilizou-se a informação disponível para a combinação fixa de Glecaprevir e Pibrentasvir (Mavyret®), medicamento usado no tratamento do vírus da hepatite C crónica. As simulações foram realizadas utilizando o software R e seu pacote Shiny. O R é uma linguagem para análise de dados de computação estatística e gráfica. A população simulada no modelo foi agrupada de acordo com as covariáveis similares, sendo simulados 1000 indivíduos por cenário. O relatório de submissão da FDA do Mavyret® foi usado como referência na modelação farmacocinética populacional. Neste relatório encontra-se descrito o modelo farmacocinético populacional desenvolvido, com base nos estudos clínicos realizados para o medicamento. No modelo descrito, foram identificadas diferentes covariáveis. O modelo descrito foi então implementado no software R e o impacto das covariáveis foi estudado com a aplicação Shiny. A população observada foi categorizada em diferentes grupos, tais como doentes tratados com Glecaprevir / Pibrentasvir com compromisso renal e doentes com compromisso renal e cirrose. Foram criados modelos individuais para cada um dos grupos e a comparação entre cada grupo e seus perfis de concentração-tempo foi realizada pelo uso do navegador R e Shiny, onde a atualização nos resultados pode ser vista automaticamente com a alteração em qualquer da covariável ou da variável. Para os diferentes modelos finais incorporados no software e para a população simulada, foram calculados os parâmetros farmacocinéticos AUC e Cmax para posterior análise estatística descritiva. Apesar da implementação dos modelos farmacocinéticos populacionais ter sido realizada em R e Shiny, e os dados terem sido simulados para os diferentes cenários populacionais, a aplicação de metodologias farmacoeconómicas não foram realizadas.
Pharmacoeconomics is the discipline concerned with optimal allocation of resources to maximize population health from the use of medicines. Given that resources for health care are finite, economic evaluation involves estimation of the opportunity cost, that is, the marginal benefits forgone as a result of displacing existing treatments or services to fund new medicines. The purpose of this study is to use tools in pharmacoeconomic analysis for the examination of the positive and adverse impact of the fixed dose combination of Glecaprevir and Pibrentasvir (Mavyret®), used to treat chronic hepatitis C virus. In order to examine the effects in pharmacoeconomics analysis, a population pharmacokinetic model was developed using R software and its package Shiny, where R is a language for data analysis of statistical computing and graphics. The population simulated in the model was grouped according to the similar covariates with the number (n) of 1000. FDA submission report for Mavyret® was used as reference regarding population pharmacokinetics modelling, developed based on the clinical studies performed for the drug product. In the described model, different covariates were identified. The described model was implemented in the R software and the impact of covariates wwas studied with Shiny application. The population observed was categorized in different groups such as patients treated with Glecaprevir/Pibrentasvir having renal impairment and patients with renal impairment and Cirrhosis. Individual models were created for each of the groups and the comparison between each group and their concentration-time profiles was observed that was made easier by the use of R and Shiny web browser where the update in results can be seen spontaneously with the change in any of the covariate or the variable. Different final models were produced and for the simulated population, the pharmacokinetic parameters AUC and Cmax were calculated for descriptive statistical analysis. Despite the implementation of population pharmacokinetics models has been accomplished in R and Shiny, and data has been simulated for different population scenarios, pharmacoeconomic modelling and application of pharmacoeconomic methodologies was not practised.

碩士
中原大學
建築研究所
90
After the industry revolution ,the human being always hold a concept of environment that man will triumph over nature for a long time .And it makes the main stream value of economic trading system keep overriding the care of ecological environment. In Taiwan,our economics grows rapidly and the industrial business property are developed vigorous from the beginning of R.O.C.. Under the background of the whole times ,the changes of ecological environment also replaced from nature to artificial, and developed behavior replaced the current environment ,the pounce on nature again after being beaten off step by step .People still haven’t be awaken by the importance of paragenesis ecological until now. The research is focus on the 1250 hectare land development case of Pu-Yu plan and the community planner system in Hsin-Chu county. They use the concept of gather village to do a hypothesize operation in specified extent and they try to find a starting point from the original power of community construction to connect with large scale of work out system. Different from the past government plan assignment, they expect to build a plan system based on the life of resident and paragenesis ecological under the current background construction. The research background are mainly divided into three premise of research to do a deep analysis : residential life, paragenesis ecological and the truely locality. The first part of research contents is using the research one himself to record local ”people” pattern in practical participated observation way and try to provide the imagination of large scale plan construction to build the follow-up operation system .The second part is using the hypothesized plan way to operate the countryside land by the station of current Pu-Yu plan. Furthermore, they use the community operation mode by community planner to discuss the real practical possibilities. The final submission part is bring tactical operation method into the whole plan and develop the ecological environment value of new era emphatically ,creating a paragenesis ecological by nature ecology and humane social system.

碩士
國立臺南大學
經營與管理學系科技管理碩士在職專班
102
This research is the first time to analyze how to build and design the new shopping centers with application service design. To understanding the position of the department store and the market segmentation. About previous research，most document indicate the practice of customer satisfaction survey and customer relationship management and goods preferences loyalty is contribute to help the company growing ,and realize the customer demand. To be the strategic direction to improve the company. Different from the past, this research focused primarily on searching the appropriate position and customer demand through the application of the service design in the initial stage. And then to conduct the plan of the tactics. This research is to investigate the company how to plan and find the position of the new shopping centers with service design in a single case study of SHIN KONG MITSUKOSH. To help the company find the clearly position, and import the new concept to bring about the market segmentation. At last, this research probe the theory and the procedure of service design and put forward the practice recommendations how the company building new centers with service design.

碩士
國立中興大學
農藝學系所
99
Use the drip irrigation system to carry out fertilizer, because fertilizer applied directly in the soil near plant roots, thus improving fertilizer use efficiently. This examination used fertilizer application through drip irrigation system to carry out the fertilizer experiment, for determining effects of different application quantities on the agronomic characters and chemical compositions of tea leaves. The experiment was located in Yuchr Branch, Tea Research and Extension Station, all treatments were arranged in a randomized complete block design, There were four treatments, which were suggested fertilization application ( N:P2O5:K2O = 70:30:40 kg / ha , T1 ) , 2 / 3 suggested fertilization application ( N:P2O5:K2O = 47:20:27 kg / ha, T2 ) , 1 / 3 suggested fertilization application (N:P2O5:K2O = 23:10:13 kg / ha, T3 ) , and non-application ( T0 ) . All treatments were replicated four times. Tea were one-year-old cutting saplings after plantation. The pH, electrical conductivity of saturation extract, and the concentrations of organic matter, total nitrogen, available phosphorus, exchangeable potassium, exchangeable calcium, and exchangeable magnesium of the soil were determined. There were four tea plucking seasons, which were in autumn, winter, spring and summer. The 100 shoots weight, internode diameter, internode length, leaf length, leaf width, leaf area, leaf thickness of the tea leaves agronomic characters were investigated, and the concentrations of total nitrogen, phosphorus, potassium, calcium, magnesium, soluble solids, total polyphenols, total catechins, caffeine, total free amino acid of tea leaves were analyzed. To analyze the soil chemical properties showing that all the experimental plot had no significant difference before treatment, this indicated the soil of experimental plot were sufficient homogeneity to be tested. After the treatment showing that the soil chemical properties will change in different degrees, only the pH and exchangeable calcium of the surface soil after treatment showed significant difference, but the difference is inconsistent with treatment. The investigative results of tea leaves agronomic characters showed T1 and T2 treatments in the second and third leaf of the leaf length, leaf width, petiole length and leaf area are the highest performance, this indicated fertilizers have significant effect on tea growth. All the nutrient contents of tea leaves showed no significant difference, this indicated fertilizer treatment had little effect on nitrogen, phosphorus, potassium, calcium and magnesium contents of tea leaves. To analyze the chemical constituents of tea, only the caffeine contents of spring tea had significant difference, T1 and T2 treatment showed the highest caffeine content, the other soluble components, total polyphenols, total catechins, total free amino acids in all tea seasons had no significant difference. Based on the above, reduced by one-third the amount of fertilizer, for agronomic characters, nutrient and chemical composition of tea did not showed decreasing trend, indicating use drip fertigation, fertilizer usage can be reduced by one-third.

碩士
華梵大學
工業設計學系碩士班
101
The many Taiwanese Temple Street, the current situation facing the the commercial recession and streetscape withered, Taiwan monuments temples weight accounted heaviest up Important characterization of civil society culture. Temple monuments retained an exhibit of the development of human society and culture activities, Specifically reflect each local Lifestyle of each era. Cultural preservation and cultural and creative experience for the demands of the Temple recycling Will form a new business model, In addition to the commercial economic redevelopment, More experiential way, A nostalgic trip story To visit the Temple of the people have the opportunity to re-simulate the life situations of the past grandparents. In this study, the selected Shin-Chuang, Taipei County, Temple Street, As actions to navigate the design area. The literature review and data collection based on the theme target Aggregate interrelated literature analysis, Through the induction of the literature, Analysis of relevant case Through fieldwork and other ways, Explore the cultural characteristics of the new Zhuang Temple Street, And the angle of the user to understand, Demand for tour guides, What's navigation allows visitors it deepened their impression of the Temple with the understanding, as a reference for the design, Temple particular style imagery development action navigation system. Help culture and enhance the value can be passed along.

碩士
國立東華大學
自然資源與環境學系
100
Taiwan’s mining site vegetation restoration technology has been developed to a certain standard, but the effectiveness and schedule of vegetation restoration in the post-mine surface of mining sites currently has no effective set of monitoring models to follow. The use of traditional manpower survey will require extensive manpower and time, and large-scale and comprehensive investigation cannot be done. In order to gain an insight into the effectiveness of vegetation restoration in the mining site, this study analyzed the Hsin-Chen Shan Quarry in Asia. A cement corporation engaged in long-term vegetation greening was the study subject. Through the satellite remote sensing image analysis technique, the results of restoration vegetation restoration in the mining site were quantified. Through the QuickBird satellite images taken in 2009, coupled with Global Digital Elevation Model (GDEM) data and aerial photos, actual mining locations over the years were compared, and the Normalized Difference Vegetation Index (NDVI) value differences of the virgin forest by the vegetation restoration area and mining site of Hsin-Chen Shan Quarry were analyzed. Additionally, the restoration results were quantified, thereby evaluating the progress of vegetation restoration in the mining site and serving as a reference for the continuous monitoring of vegetation recovery in mining sites in the future. The results show that the main factor affecting the quarry NDVI value changes is the growth of vegetation, which can obscure rocks. In addition, the virgin forest vegetation is more dense and less subject to destruction. The NDVI standard deviation is smaller, which is significantly different from the NDVI standard deviation in the restoration area. On the other hand, the NDVI value in the restoration area tends to increase with time. That is, areas with longer restoration time have higher NDVI mean values. The regression analysis results show that when the mining site’s vegetation restoration time exceeded 22 years, the NDVI value also reached the greening level of the virgin forest. In the future, remote sensing image analysis techniques with such advantages as a wide range of investigative capacity and real-time monitoring can be adopted to reduce the difficulty of monitoring vegetation restoration in mining sites. In this study, through field planting year, combined with the NDVI analysis method, vegetation greening was effectively obtained. By applying this model in mining sites to conduct long-term vegetation monitoring, the mining site vegetation restoration related explorations can effectively keep pace with the times.

碩士
國立宜蘭大學
土木工程學系碩士班
103
In Taiwan, the methods currently used for flooding area estimation are mainly based on the after-survey of the flood mark. These methods can not reflect the real-time situation of the flooding area, and hence can’t able to provide immediate information for damage control. In this study, we set up automatic water stage monitoring system at the Mei-Fu region where is selected as the target area of the project due to frequent flood history of this area. We set many model and used the ANFIS(Adaptive Network-Based Fuzzy Inference System) to training the rainfall data of the flood events and the data from water stage, then establish the relationship between rainfall and flooding depth. Used the Coefficient of Efficiency(CE), Peak Error(PE) and Time Shift Error(TSE), maximum reached time of water level to verify the results. To find the best result model. The results indicated that the proposed model can reasonably simulate change trend of water level.

碩士
南華大學
美學與藝術管理研究所
95
Many countries take education as a decisive factor of their national competitiveness, hence in the 21st century; education reforms are taking place in many societies. Art education in Taiwan is facing challenges while an overall education reform is undertaking. Having been ignored for decades, the status of the art education has been changed in Taiwan since the implementation of the "Integrated Curriculum in Grade 1－9 ". 　 　　Thanks to the efforts of many art educators, art is possible to be taught during our daily activities through our interactions with the environment. This study explores the potential of the Situated Learning in art education by trying several teaching methods in the Tung-Shin Elementary School. The conclusion is made based on teachers'' observation and interviews with students. This study also discusses the effects of integrating Situated Learning into art education. 　 　　According to the results this study has found, educators might include the learning methods that students take charge instead of teachers, authentic situation emulation, high degree student participation, a learning environment concerning both humanity and nature , flexible curriculums and learning programs aspired by daily life. Some other strategies are also recommended: application of cognitive apprenticeship, collaborative learning and cell group learning emphasis, application of multimedia, learning environment enlargement and art appreciation programs among others. The strategies and principles this study has established, hopefully, are able to provide art teachers with a variety of options in art education.

碩士
國立屏東科技大學
農企業管理系所
106
For the past few years, the industry in Taiwan had already be changed from the farming industry-oriented to the service industry-oriented; moreover, the agricultural work needs less and less manpower because of mechanized promotion. So the Agricultural environment was pushed out of business. Therefore, the administration provides the revitalization plans in order to find the solution way for the current and even the future of the agricultural. In one of the revitalization plans, the administration works out the transition strategy to help the poor agricultural and animal husbandry successful to make the transition to an educational leisure farm. Even though, in some of the useless transition strategy and the weakest effectiveness of its implementation also get failing result. Based on the above, for purpose of formulating the leisure farmer’s administration strategy, to collect the relative literature of industry strategy and discuss the analysis of the method and the strategy of mode is necessary. In order to make the suggestion for the leisure agricultural how to develop persistently. By interview, I am going to understand the competitive advantage for domestic Leisure agriculture, then analyze the opportunity and threat of External environment for overall economy, also find out the management strategy via analyze internal strength and weakness. So, how to improve the management strategy and in response to trends for leisure agriculture is the key point for this research. Finally, I also expect to find out the sustainable management strategy and Zhao Feng farmer take it as a reference source.

碩士
中國文化大學
青少年兒童福利研究所
98
This research adopted the method of action research and the samples are twelve new female immigrants of a life adjustment class at one Women and Children Service Center. The purposes of this research were to explore the new female immigrants’ comprehension and learning of their development, as for creative drama incorporating a life adjustment class, and to plan eight units in one course, and to assess the scheme at its effect of implementation and the influence on the new female immigrants’ development of life adjustment. Through taking action research to ponder over the problems and solution in the process of teaching and learning. The results of this study also provide a reference for arrangement the courses of life adjustment of new female immigrants. The tools for collecting and analyzing the materials through the participating observation, the teaching through drama course given on one women and children service center, interview record, record of every class of research and the performance of achievements. According to the findings of this action research has obvious effect on Language Learning Programs of the New Female Immigrants.