xvi, 167 p. A print copy of this thesis is available through the UO Libraries. Search the library catalog for the location and call number. / The dynamic scripting reinforcement learning algorithm can be extended to improve the speed, effectiveness, and accessibility of learning in modern computer games without sacrificing computational efficiency. This dissertation describes three specific enhancements to the dynamic scripting algorithm that improve learning behavior and flexibility while imposing a minimal computational cost: (1) a flexible, stand alone version of dynamic scripting that allows for hierarchical dynamic scripting, (2) a method of using automatic state abstraction to increase the context sensitivity of the algorithm, and (3) an integration of this algorithm with an existing hierarchical behavior modeling architecture. The extended dynamic scripting algorithm is then examined in the three different contexts. The first results reflect a preliminary investigation based on two abstract real-time strategy games. The second set of results comes from a number of abstract tactical decision games, designed to demonstrate the strengths and weaknesses of extended dynamic scripting. The third set of results is generated by a series of experiments in the context of the commercial computer role-playing game Neverwinter Nights demonstrating the capabilities of the algorithm in an actual game. To conclude, a number of future research directions for investigating the effectiveness of extended dynamic scripting are described. / Adviser: Arthur Farley

In this paper research about dynamic scripting and the top culling difficulty scaling enhancement in the game Neverwinter Nights is investigated. A comparison between both a static and a dynamic opponent is made. The human opinion about dynamic scripting is also highlighted. To get an understanding of what the players think about and how they approach an opponent, two experiments were made. One where tests are made on a static opponent and a dynamic opponent, then a second where differences in behavior of the dynamic opponent using top culling and an ordinary dynamic opponent are analyzed. Results from the first test shows the static opponent is more preferable whereas the dynamic opponent using top culling is preferred in the second experiment. Since comparing the two experiments the results are ambiguous. The conclusion is that further investigation is needed in order to answer the question if human players prefer static or dynamic opponents when playing computer games. / I detta arbete undersöks tekniken "Dynamic Scripting" och en metod för att skala svårigheten hos motståndare, kallad "Top Culling". Detta har testats i spelet "Neverwinter Nights". En jämförelse mellan en statisk och en dynamisk motståndare har gjorts där den mänskliga synen på dynamic scripting är en huvudfaktor. För att få förståelse hur spelare tänker och bemöter olika motstånd gjordes två experiment. I ett av experimenten testas en statisk och en dynamisk motståndare. I ett annat experiment görs en analys av skillnaderna i beteende mellan en dynamisk motståndare med svårighetsskalning och en dynamisk motståndare utan. Det första experimentet gav resultat som visar att den statiska motståndaren föredras medan i det andra experimentet föredras den dynamiska motståndaren som skalade sin svårighetsgrad. Slutsatsen är att vidare undersökning krävs för att kunna besvara frågan huruvida en spelare hellre vill möta en statisk eller en dynamisk motståndare.

?rea de concentra??o: Educa??o e Tecnologias aplicadas em Institui??es Educacionais. / Submitted by Jos? Henrique Henrique (jose.neves@ufvjm.edu.br) on 2018-04-12T14:17:26Z No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) julio_cesar_costa_silva.pdf: 3442052 bytes, checksum: 418632e21262716d31a67fd4b2f368f5 (MD5) / Approved for entry into archive by Rodrigo Martins Cruz (rodrigo.cruz@ufvjm.edu.br) on 2018-04-23T16:45:43Z (GMT) No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) julio_cesar_costa_silva.pdf: 3442052 bytes, checksum: 418632e21262716d31a67fd4b2f368f5 (MD5) / Made available in DSpace on 2018-04-23T16:45:43Z (GMT). No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) julio_cesar_costa_silva.pdf: 3442052 bytes, checksum: 418632e21262716d31a67fd4b2f368f5 (MD5) Previous issue date: 2017 / Uma das formas de se gerar conte?do adaptado ao estudante passa, primeiro, pela detec??o dos Estilos de Aprendizagem (EA). A teoria dos EA presume que cada aluno tem caracter?sticas pr?prias que o distingue dos demais. A partir dos EA, o Sistema Adaptativo e Inteligente para Educa??o (SAIE) de Dor?a foi idealizado. Seu trabalho objetiva apresentar uma solu??o estoc?stica para provimento de adaptatividade e customiza??o de Sistemas Educacionais por meio da modelagem probabil?stica dos EA. Em s?ntese, seu SAIE visa modelar o estudante, coletando e atualizando seus dados, de forma a descobrir seu EA. Com este fim, o sistema, durante suas itera??es, submete o aluno a avalia??es e, caso as notas sejam insatisfat?rias, o sistema realiza a atualiza??o do modelo do estudante (ME) por meio do Aprendizado por Refor?o (AR). Contudo, AR ? considerada uma t?cnica lenta de aprendizado que demanda muito tempo para ajustar o elemento a ser otimizado. Por sua vez, a t?cnica Dynamic Scripting (DS), uma varia??o da t?cnica de AR, apresenta alta velocidade de converg?ncia, mesmo em ambientes din?micos. DS ? popularmente utilizada na IA de Jogos e consiste em um conjunto de Regras sobre um dom?nio, estruturadas por uma condi??o e uma a??o. Sua forma de aprendizagem atrela um peso a cada regra, o qual determina a qualidade da regra, frente ? sua condi??o, e uma probabilidade da mesma ser aplicada. A condi??o de uma regra ? a representa??o de uma situa??o poss?vel no sistema, e sua a??o ? a interven??o gerada no sistema durante a sua aplica??o. Este trabalho prop?e o aperfei?oamento do SAIE citado, utilizando uma adapta??o do DS, com os objetivos de acelerar a converg?ncia do sistema, reduzir os Problemas de Aprendizagem (PA) e aumentar a nota do estudante. Adicionalmente, devido a caracter?stica din?mica do DS, este trabalho realiza experimentos em situa??es em que o EA Real (EAr) dos alunos variam ao longo do processo de ensino/aprendizagem. A pesquisa parte da elabora??o das regras e implementa??o da estrutura do DS, avan?ando para a substitui??o do m?dulo de AR pelo DS no SAIE de Dor?a. Realizaram-se 30 testes para cada uma das 16 Combina??es de EA (CEA), 16*30 testes para cada uma das 4 abordagens: Dor?a-Est?tico, Dor?a-Din?mico, DS-Est?tico e DS-Din?mico. Nos testes din?micos, modificou-se o EAr a cada 150 intera??es, de forma que ap?s 300 intera??es, o sistema deve convergir para uma CEA oposta ? inicial. Resultados preliminares, em compara??o ? abordagem da literatura, apresentaram uma redu??o m?dia nos PA de 35.8% para os testes din?micos e de 54.1% para os testes est?ticos. Quando o EA Probabil?stico (EAp) inicial ? exatamente igual ao EAr, verificou-se que a abordagem proposta apresentou em m?dia 6 erros na atualiza??o do ME, enquanto a abordagem da literatura apresentou, em m?dia, 23 erros. Verificou-se, portanto, que, preliminarmente, a proposta obteve resultados promissores. / Disserta??o (Mestrado Profissional) ? Programa de P?s-Gradua??o em Educa??o, Universidade Federal dos Vales do Jequitinhonha e Mucuri, 2017. / One of the ways to generate the content adapted to the student passes, first, by the detection of Learning Styles (LS). The LS theory assumes that every time you have distances. From the LS, the Adaptive and Intelligent System for Education (AISE) of Dor?a was idealized. His work aims to present a stochastic solution for the provision of adaptability and customization of Educational Systems through the probabilistic modeling of LS. In summary, your AISE visa model, offering and updating your data, in order to discover your LS. To this end, the system, during its iterations, submits the student to the evaluation and, in case of notes and dissatisfactions, the system performs an updating of the student model (ME) through Reinforcement Learning (RL). However, RL is a slow learning technique that requires a lot of time to adjust the element to be optimized. In turn, a Dynamic Scripting (DS) technique, a variation of the RL technique, presents a high speed of convergence, even in dynamic environments. DS is popularly used in Artificial Intelligence of Games and consists of a set of Rules on a domain, structured by a condition and an action. Its form of learning brings a weight to each rule, which determines a quality of the rule, in front of its condition, and a probability of the same company. The condition of a rule is a representation of a good situation, and its action is an intervention generated without system during its application. This work proposes the improvement of the SAIE mentioned, the use of an adaptation of the DS, with the objectives of accelerating the convergence of the system, reduce the Learning Problems (PA) and increase student grade. In addition, due to the dynamic nature of the DS, this work performs tasks in situations in which students? real LS (LSr) vary throughout the teaching / learning process. A research of elaboration of the rules and implementation of the structure of the DS, advancing to the substitution of the RL module by the DS without AISE of Dor?a. A total of 30 tests were performed for each of the 16 AE combinations (CEA), 16 * 30 testicles for each of the 4 approaches: Dorca-Static, Dynamic Doric, DS-Static and DS-Dynamic. In the dynamic tests, the LSr was modified every 150 interactions, so that after 300 interactions, the system must converge to a CEA opposite to the initial one. Preliminary results, in literature comparison, presented a mean reduction in BP of 35.8 % for dynamic tests and 54.1 % for static tests. When the initial Probabilistic LS (LSp) is exactly the same as the LSr, it was verified that the proposed approach presented on average 6 errors in the updating of the ME, while a literature approach presented, on average, 23 errors. It was therefore found that a proposal had obtained promising results in the first place.