Making information stored in databases more accessible to users inexperienced in structured query language (SQL) by converting natural language to SQL queries has long been a prominent research area in both the database and natural language processing (NLP) communities. There have been numerous approaches proposed for this task, such as encoder-decoder frameworks, semantic grammars, and more recently with the use of large language models (LLMs). When training LLMs to successfully generate SQL queries from natural language questions there are three notable methods used, pretraining, transfer learning and in-context learning (ICL). ICL is particularly advantageous in scenarios where the hardware at hand is limited, time is of concern and large amounts of task specific labled data is nonexistent. This study seeks to evaluate two strategies in ICL, namely zero-shot and few-shot scenarios using the Mistral-7B-Instruct LLM. Evaluation of the few-shot scenarios was conducted using two techniques, random selection and Jaccard Similarity. The zero-shot scenarios served as a baseline for the few-shot scenarios to overcome, which ended as anticipated, with the few-shot scenarios using Jaccard similarity outperforming the other two methods, followed by few-shot scenarios using random selection coming in at second best, and the zero-shot scenarios performing the worst. Evaluation results acquired based on execution accuracy and exact matching accuracy confirm that leveraging similarity in demonstrating examples when prompting the LLM will enhance the models knowledge about the database schema and table names which is used during the inference phase leadning to more accurately generated SQL queries than leveraging diversity in demonstrating examples.

This thesis explores the application of prompt engineering combined with human-AI interaction (HAII) to make text-to-image (TTI) models more accessible and intuitive for non-expert users. The thesis research focuses on developing an application with an intuitive interface that enables users to generate images without extensive knowledge of prompt engineering. A pre-post study was conducted to evaluate the application, demonstrating significant improvements in user satisfaction and ease of use. The findings suggest that such tailored interfaces can make AI technologies more accessible, empowering users to engage creatively with minimal technical barriers. This study contributes to the fields of Media technology and AI by showcasing how simplifying prompt engineering can enhance the accessibility of generative AI tools. / Detta examensarbete utforskar tillämpningen av prompt engineering i kombination med human-AI interaction för att göra text-till-bild modeller mer tillgängliga och intuitiva för icke-experter. Forskningen för examensarbetet fokuseras på att utveckla en applikation med ett intuitivt gränssnitt som gör det möjligt för användare att generera bilder utan omfattande kunskaper om prompt engineering. En före-efter-studie genomfördes för att utvärdera applikationen, vilket visade på en tydlig ökning i användarnöjdhet och användarvänlighet. Utfallet från studien tyder på att skräddarsydda gränssnitt kan göra AI-tekniken mer tillgänglig, och göra det möjligt för användare att nyttja det kreativa skapandet med minimerade tekniska hinder. Den här studien bidrar till områdena avmedieteknik och AI genom att demonstrera hur prompt engineering kan förenklas vilket kan förbättra tillgängligheten av AI-verktyg.

Ion beam therapy is an innovative radiotherapy technique using mainly carbon ion and proton irradiations. Its aim is to improve the current treatment modalities. Because of the sharpness of the dose distributions, a control of the dose if possible in real time is highly desirable. A possibility is to detect the prompt gamma rays emitted subsequently to the nuclear fragmentations occurring during the treatment of the patient. In a first time two different Compton cameras (double and single scattering) have been optimised by means of Monte Carlo simulations. The response of the camera to a photon point source with a realistic energy spectrum was studied. Then, the response of the camera to the irradiation of a water phantom by a proton beam was simulated. It was first compared with measurement performed with small-size detectors. Then, using the previous measurements, we evaluated the counting rates expected in clinical conditions. In the current set-up of the camera, these counting rates are pretty high. Pile up and random coincidences will be problematic. Finally we demonstrate that the detection system is capable to detect a longitudinal shift in the Bragg peak of +or- 5 mm, even with the current reconstruction algorithm.

Au cours du traitement d'une tumeur avec des ions légers, la position du pic de Bragg doit être connue avec précision. Une fraction importante des ions incidents subissent des collisions nucléaires avec les noyaux cibles conduisant à l'émission de particules promptes qui peuvent être porteuses d'informations sur le parcours des ions. Ce travail, qui se concentre sur les gamma prompts, montre que le rendement en profondeur de ces émissions est fortement corrélé au parcours des ions et que les taux de comptage mesurés permettent d'envisager un système d'imagerie réaliste, fonctionnant en temps réel. Dans ce but, nous avons réalisé des expériences au GANIL et au GSI avec un détecteur collimaté placé perpendiculairement à l'axe du faisceau et la technique du temps de vol a été utilisée pour réduire le bruit de fond induit par les neutrons et les particules chargées. Des simulations Geant4 ont été réalisées pour concevoir le dispositif expérimental et interpréter les données. Un accord qualitatif entre les simulations et l'expérience est observé pour la quantité d'énergie déposée dans le détecteur et pour la forme du spectre de temps de vol. Cependant, des divergences apparaissent pour le rendement de gamma prompts et la distribution en profondeur des gamma détectés. Ces divergences sont discutées, principalement en termes de modèles de physique nucléaire qui doivent être améliorés. Après avoir sélectionné les modèles physiques offrant les simulations les plus en accord avec les mesures, des études concernant les lieux d'émissions des gamma prompts et l'influence de la diffusion dans la cible ont été réalisés afin de déterminer l'impact sur la corrélation avec le parcours des ions / During the treatment of a tumor with light ions, the Bragg peak location must be accurately known. A significant fraction of the incident ions undergo nuclear collisions with the target nuclei leading to the prompt emission of particles which may carry information on the ion path. This work, which focuses on prompt gamma, shows that the depth profile of these emissions is highly correlated to the ions path and the measured counting rates allow to consider a realistic imaging system, operating in real time. For that purpose, we performed experiments at GANIL and at GSI with a collimated detector placed perpendicular to the beam axis and the time of flight technique was used in order to reduce the noise induced by neutrons and charged particles. Geant4 simulations were performed for the experimental design and data interpretation. A qualitative agreement between simulations and experiment is observed for the amount of energy deposited in the detector and the shape of the time of flight spectrum. However, discrepancies appear for the prompt gamma yield and the depth distribution of gamma detected. These discrepancies are discussed, mainly in terms of nuclear physics models that must be improved. After selecting the physical models which lead to the best agreement between simulations and measurements, studies on the location of prompt gamma emission and on the influence of diffusion in the target were performed to determine the impact on the correlation with the ion path

Les noyaux super lourds offrent la possibilité d’étudier la structure nucléaire à trois limites simultanément: en charge Z, spin I et énergie d’excitation E∗. Ces noyaux existent seulement grâce à une barrière de fission créée par les effets de couche. Il est donc important de déterminer cette barrière de fission et sa dépendance en spin Bf(I), qui nous renseigne sur l’énergie de couche Eshell(I). Les théories prédisent des valeurs différentes pour la hauteur de la barrière de fission, allant de Bf(I = 0) = 6.8 MeV dans un modèle macro-microscopique à 8.7 MeV pour des calculs de la théorie de la fonctionnelle de la densité utilisant l’interaction Gogny ou Skyrme. Une mesure de Bf fournit donc un test des théories.Pour étudier la barrière de fission, la méthode établie est de mesurer, par réaction de transfert, l’augmentation de la fission avec l’énergie d’excitation, caractérisée par le rapport des largeurs de décroissance Γfission/Γtotal,. Cependant, pour les éléments lourds comme 254No, il n’existe pas de cible appropriée pour une réaction de transfert. Il faut s’en remettre à un rapport de largeur de décroissance complémentaire: Γγ/Γfission et sa dépendance en spin, déduite de la distribution d’entrée (I, E∗).Des mesures de la multiplicité et l’énergie totale des rayons γ de254No ont été faites aux énergies de faisceau 219 et 223 MeV pour la réaction 208Pb(48Ca,2n) à ATLAS (Argonne Tandem Linac Accelerator System). Les rayons γ du 254No ont été détectés par le multi-détecteur Gammasphere utilisé comme calorimètre – et aussi comme détecteur de rayons γ de haute résolution. Les coïncidences avec les résidus d’évaporation au plan focal du Fragment Mass Analyzer ont permis de séparer les rayons γ du 254No de ceux issus de la fission, qui sont > 10^6 fois plus intenses. De ces mesures, la distribution d’entrée – c’est-à-dire la distribution initiale en I et E∗ – est reconstruite. Chaque point (I,E∗) de la distribution d’entrée est un point où la décroissance γ l’a emporté sur la fission, et ainsi, contient une information sur la barrière de fission.La distribution d’entrée mesurée montre une augmentation du spin maximal et de l’énergie d’excitation entre les énergies de faisceau 219 et 223 MeV. La distribution présente une saturation de E∗ à hauts spins. Cette saturation est attribuée au fait que, lorsque E∗ augmente au-dessus de la barrière, Γfission domine rapidement. Il en résulte une troncation de la distribution d’entrée à haute énergie qui permet la détermination de la hauteur de la barrière de fission.La mesure expérimentale de la distribution d’entrée est également comparée avec des distributions d’entrée calculées par des simulations de cascades de décroissance qui prennent en compte le processus de formation du noyau, incluant la capture et la survie, en fonction de E∗ et I. Dans ce travail, nous avons utilisé les codes KEWPIE2 et NRV pour simuler les distributions d’entrée. / Super heavy nuclei provide opportunities to study nuclear structure near three simultaneous limits: in charge Z, spin I and excitation energy E∗. These nuclei exist only because of a fission barrier, created by shell effects. It is therefore important to determine the fission barrier and its spin dependence Bf(I), which gives information on the shell energy Eshell(I). Theoretical calculations predict different fission barrier heights from Bf(I = 0) = 6.8 MeV for a macro-microscopic model to 8.7 MeV for Density Functional Theory calculations using the Gogny or Skyrme interactions. Hence, a measurement of Bf provides a test for theories.To investigate the fission barrier, an established method is to measure the rise of fission with excitation energy, characterized by the ratio of decay widths Γfission/Γtotal, using transfer reactions. However, for heavy elements such as 254No, there is no suitable target for a transfer reaction. We therefore rely on the complementary decay widths ratio Γγ/Γfission and its spin dependence, deduced from the entry distribution (I, E∗).Measurements of the gamma-ray multiplicity and total energy for 254No have been performed with beam energies of 219 and 223 MeV in the reaction 208Pb(48Ca,2n) at ATLAS (Argonne Tandem Linac Accelerator System). The 254No gamma rays were detected using the Gammasphere array as a calorimeter – as well as the usual high resolution γ-ray detector. Coincidences with evaporation residues at the Fragment Mass Analyzer focal plane separated 254No gamma rays from those from fission fragments, which are > 10^6 more intense. From this measurement, the entry distribution – i.e. the initial distribution of I and E∗ – is constructed. Each point (I,E∗) of the entry distribution is a point where gamma decay wins over fission and, therefore, gives information on the fission barrier.The measured entry distributions show an increase in the maximum spin and excitation energy from 219 to 223 MeV of beam energy. The distributions show a saturation of E∗ for high spins. The saturation is attributed to the fact that, as E∗ increases above the saddle, Γfission rapidly dominates. The resulting truncation of the entry distribution at high E∗ allows a determination of the fission barrier height.The experimental entry distributions are also compared with entry distributions calculated with decay cascade codes which take into account the full nucleus formation process, including the capture process and the subsequent survival probability as a function of E∗ and I. We used the KEWPIE2 and NRV codes to simulate the entry distribution.

SOFIA (Studies On FIssion with Aladin - Études de fission avec ALADIN) est un projet expérimental qui se propose de mesurer systématiquement les rendements isotopiques de fission, ainsi que l'énergie cinétique totale des fragments, pour une large gamme de noyaux fissionnants. Le travail de thèse présenté ici prend part au projet SOFIA, et a pour objet l'étude de la fission de noyaux dans la région des actinides : ²³⁴U, ²³⁵U, ²³⁷Np et ²³⁸Np.L'expérience SOFIA est menée au GSI, un accélérateur d'ions lourds situé à Darmstadt en Allemagne. Cette installation fournit un faisceau relativiste et intense d'uranium-238. Par une réaction de fragmentation de ce faisceau primaire, il est créé un faisceau secondaire contenant une large variété d'ions, dont, pour certains, la fission va être étudiée. Les ions de ce faisceau secondaire sont triés et identifiés dans le séparateur de fragments FRS, un spectromètre de recul de haute résolution qui est réglé pour sélectionner les ions d'intérêt.Ensuite, les ions fissiles sélectionnés continuent leur vol jusqu'à la Cave-C, une aire expérimentale où l'expérience de fission elle-même a lieu. À l'entrée de la cave, l'ion du faisceau secondaire est excité par interaction Coulombienne alors qu'il traverse une cible; les voies de dé-excitation possibles incluent la fission de basse énergie. Lors d'une fission, les deux fragments sont émis dans un cône étroit dans le référentiel du laboratoire, dû à l'impulsion relativiste qui leur est transmise par le système fissionnant. Un spectromètre de recul complet a été développé par la collaboration SOFIA autour du dipôle existant ALADIN. L'identification des fragments est réalisée par des mesures de perte d'énergie, de temps de vol et de déviation dans l'aimant. Les deux fragments de fission sont identifiés simultanément et complètement (en masse et en charge). Ce document présente l'analyse menée pour (1) l'identification du système fissionnant, (2) l'identification des deux fragments de fission - évènements par évènements -, et (3) l'extraction des observables de fission : rendements, énergie cinétique totale, multiplicité total en neutrons prompts. Ces résultats concernant les actinides sont discutés, et l'ensemble des données extraites est fourni. / SOFIA (Studies On FIssion with Aladin) is an experimental project which aims at systematically measuring the fission fragments' isotopic yields as well as their total kinetic energy, for a wide variety of fissioning nuclei. The PhD work presented in this dissertation takes part in the SOFIA project, and covers the fission of nuclei in the region of the actinides : ²³⁴U, ²³⁵U, ²³⁷Np and ²³⁸Np.The experiment is led at the heavy-ion accelerator GSI in Darmstadt, Germany. This facility provides intense relativistic primary beam of 238U. A fragmentation reaction of the primary beam permits to create a secondary beam of radioactive ions, some of which the fission is studied. The ions of the secondary beam are sorted and identified through the FR-S (FRagment Separator), a high resolution recoil spectrometer which is tuned to select the ions of interest.The selected - fissile - ions then fly further to Cave-C, an experimental area where the fission experiment itself takes place. At the entrance of the cave, the secondary beam is excited by Coulomb interaction when flying through an target; the de-excitation process involves low-energy fission. Both fission fragments fly forward in the laboratory frame, due to the relativistic boost inferred from the fissioning nucleus.A complete recoil spectrometer has been designed and built by the SOFIA collaboration in the path of the fission fragments, around the existing ALADIN magnet. The identification of the fragments is performed by means of energy loss, time of flight and deviation in the magnet measurements. Both fission fragments are fully (in mass and charge) and simultaneously identified.This document reports on the analysis performed for (1) the identification of the fissioning system, (2) the identification of both fission fragments, on an event-by-event basis, and (3) the extraction of fission observables: yields, TKE, total prompt neutron multiplicity. These results, concerning the actinides, are discussed, and the set of data extracted is provided.

Teachers of foreign languages typically encourage their students to speak in the target language (TL) often, but due to various factors, this is not always achieved. Some reasons might include insufficient vocabulary, lack of topic knowledge, embarrassment, or simple unwillingness. Much of the existing research observes uses of the TL or native language (L1). The purpose of this study was to examine how often students use the TL in paired interactions and whether that amount has any relationship to the students' oral proficiency at the end of the course. In a SPAN 105 course at Brigham Young University, 27 students participated in this project by recording themselves during six in-class, paired interactions, after which they provided comments via questionnaires. This study was conducted using a mixed-methods approach, with both quantitative and qualitative data. The quantitative data revealed information about the time spent in the TL, as well as the time spent in the L1, and what relationship these variables had with listening comprehension and oral proficiency. The qualitative data exposed emergent findings related to TL/L1 use: helpful tools, effects of task type, pitfalls the students experienced, struggles and benefits of partnering, effects of recording, and student perceptions about the L1. The results of the study indicate a need for teacher strategies to encourage TL use in the classroom.

農漁會組織長久以來於農業發展扮演重要角色，其促進農業生產，增進農民福祉與繁榮農村經濟，對台灣早期之經濟發展貢獻許多。隨著經濟結構的變動，農會信用部面對產業結構的轉變及其他金融機構的激烈競爭，致使其經營陷入困境，經過金融重建基金處理了36家經營不善信用部，宣佈分級管理措施，及12萬農民大遊行，政府為徹底解決信用部諸多問題，於93年1月實施農業金融法，建立由行政院農業委員會一元化管理之農業金融體系，經過一連串之改革，相關財務指標顯示信用部之經營已逐漸改善中。 本研究主要探討我國農業金融體制與信用部面臨的問題，並參考日本農業金融改革之經驗，提出健全我國農業金融體制與監理之建議。 研究發現，農業金融改革後信用部之經營確實在改善中。然而，在80年代農業金融危機下遺留的問題尚未完全克服。對於我國農業金融發展，本研究從組織面、業務面與監理面進行探討，提出改革建議包括：全面檢討修訂農會法與漁會法、儘速恢復股金制、建置合併法規鼓勵合併、儘速處理經營不善之信用部並建構多元退場機制、加強農業金融體系連結與加速資訊共用平台之整合、強化對全國農業金庫與信用部之監理、落實金融監理加強實地檢查與場外監控措施、導入差異化管理與立即糾正措施等，農業金融機構有必要繼續改革，以健全農業金融體系，保障存款人權益，促進農漁村經濟發展。 / The Farmers’ and Fishermen’s Associations played an important role in the field of agricultural production. They helped agriculture develop, increased farmers’ and fishermen’s welfare, flourished the countryside and contributed a lot to the early progress of Taiwan economy. With the transition of economic structure, the whole environment became quite disadvantageous to credit departments of farmers’ and fishermens’ associations, which face the changes of the industrial structure and fierce competition from other financial institutions. With the experiences of the settlement of 36 problem credit departments by the Financial Restructuring Fund, announcement of differential supervision scheme and demonstration of 120,000 agriculturists, the government implemented the Agricultural Finance Act on 30th January 2004 and built an integrated agricultural finance system governed by the Council of Agriculture (COA) to solve many problems of credit departments. Through those reformations, the financial index showed that the condition of these credit departments has improved gradually. This study aimed to discuss those difficulties that our agricultural finance system and credit departments encountered and bring up suggestions to complete this system and the government’s supervision referring to the reformation of agricultural finance system in Japan. What our study found is that the operation of credit departments has undoubtedly improved after taking reformations to agricultural finance system in Taiwan. However, problems that the agricultural finance crisis left behind in 1980s have not been completely conquered yet. As to the prospect of our agricultural finance system, from the aspects of organizational structures, business activities and government’s supervision, we suggest the reforms include to examine and amend both the Farmers’ Association Law and the Fishermen’s Association Law from stem to stern, re-enforce paid-in capital system with all speed, draw up laws to encourage mergers, deal with problem credit departments and build up plenty selections to help them exit as soon as possible, strengthen the connection of agricultural finance system and the integration of information sharing stations, intensify our supervision towards the Agricultural Bank of Taiwan and credit departments, reinforce on-the-spot examination and off-site monitoring, bring differential supervision scheme and prompt-corrective action into practice and so on. Hence, we may achieve the goal to complete the agricultural finance system, protect the rights of depositors and prosper the rural villages’ and fishing villages’ economy.

Made available in DSpace on 2014-10-09T12:53:07Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T13:59:04Z (GMT). No. of bitstreams: 0 / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / Tese (Doutoramento) / IPEN/T / Instituto de Pesquisas Energeticas e Nucleares - IPEN/CNEN-SP / FAPESP:03/01261-0

Made available in DSpace on 2014-10-09T12:53:07Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T13:59:04Z (GMT). No. of bitstreams: 0 / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / Uma nova metodologia para a medida absoluta da fração efetiva de nêutrons atrasados βeff, baseada em técnicas de análise de ruído microscópico e no modelo de Duas- Regiões, foi desenvolvida no reator IPEN/MB-01. Diferentemente das demais técnicas, tais como o Método de Bennet Modificado, o Método do Número de Nelson e o Método da fonte de 252Cf, a principal vantagem da metodologia proposta é a obtenção de βeff de um modo puramente experimental, sem a necessidade de quaisquer outros parâmetros, sejam estes calculados ou provenientes de outros experimentos. Com a finalidade de validar este novo método, uma série de experimentos Rossi-α e Feynman-α foram realizados no reator IPEN/MB-01. De acordo com a metodologia proposta, βeff foi estimado com uma incerteza de 0.67%, a qual atende aos requisitos de precisão almejados. Além disso, o tempo de geração de nêutrons prontos , dentre outros parâmetros, também foi obtido experimentalmente via esta metodologia. Em geral, os parâmetros medidos estão em acordo com resultados provenientes de experimentos de análise de ruído macroscópico. Nas comparações teoria-experimento, os valores de βeff medidos neste trabalho mostram que a biblioteca JENDL3.3 apresenta a melhor performance (dentro de 1%). Esta concordância justifica a redução no yield de fissão do 235U proposta por Sakurai e Okajima. / Tese (Doutoramento) / IPEN/T / Instituto de Pesquisas Energeticas e Nucleares - IPEN/CNEN-SP / FAPESP:03/01261-0