Programming industrial robots for welding or part manipulation tasks is a time-consuming and complicated process, resulting in companies not able to implement robot systems and exploit their advantages. To reduce the time needed for programming, research is looking into ways to automate this process and reduce manual labour.In this thesis a concept for automating the programming process of industrial robots was investigated using EXCEL VBA and CATIA/DELMIA. It was done for an industrial grating model of varying sizes and configurations, resulting in a time reduction of 99% compared to manual creation. For this, the model was first automatically created from scratch for the required configuration and afterwards a robot motion was created fully automatically. The concept and modelling approach is described, and the automation approach detailed. Finally, the results are analysed and discussed.

Driven by recent technical advances, vehicle-based transportation systems in intralogistics are currently shifting from automated guided vehicles (AGVs) to autonomous mobile robots (AMRs). Unlike AGVs, AMRs are not bound to a physical or virtual track and autonomously determine their movements. While the increase in freedom for routing leads to improvements in terms of system flexibility and scalability, it also poses new challenges in terms of coordination and thus the high-performance routing behavior of an entire fleet. Accepted and widely used algorithms in the AGV field are often barely applicable to large fleets of free-ranging AMRs, while typical algorithms from the field of mobile robotics usually focus on different objectives. This thesis presents a novel concept for determining global routes, the so-called lanemap, that enables the synchronized movements of multiple free-ranging vehicles in arbitrary layouts without increasing calculation effort during online operations. The basic idea consists in creating a lanemap offline that provides the AMRs with a set of suggested global lanes from various starting positions to different destinations. On the one hand, the application of a lanemap lowers each AMRs’ individual degree of freedom. However, coordinating lanes sensibly beforehand enables short travel distances as well as a low probability of routing conflicts, which improves the performance of the entire AMR fleet on the other hand. Since the lanemap only provides an offline calculated global route as suggestion, the concept can be combined with any established online approach, such as a local conflict avoidance/resolution approach, or with in-depth coordination of all vehicle movements. This thesis presents a theoretical mathematical model and a practically applicable heuristic approach for the creation of a lanemap. As proof of concept, simulation experiments show that the heuristic is generally capable of creating lanemaps for all different kinds of layouts. Furthermore, the concept allows for the determination of customized routes for a specific fleet size and an anticipated transportation demand. Therefore, once system requirements are known, a beneficial set of lanemaps for typical constellations can be calculated in advance and integrated into routing as needed.

<p dir="ltr">Reinforcement learning (RL) has been a rapidly evolving field of research over the past years, enhancing developments in areas such as artificial intelligence, healthcare, and education, to name a few. Regardless of the success of RL, its inherent online learning nature presents obstacles for its real-world applications, since in many settings, online data collection with the latest learned policy can be expensive and/or dangerous (such as robotics, healthcare, and autonomous driving). This challenge has catalyzed research into offline RL, which involves reinforcement learning from previously collected static datasets, without the need for further online data collection. However, most existing offline RL methods depend on two key assumptions: unconfoundedness and positivity (also known as the full-coverage assumption), which frequently do not hold in the context of static datasets. </p><p dir="ltr">In the first part of this dissertation, we simultaneously address these two challenges by proposing a novel policy learning algorithm: PESsimistic CAusal Learning (PESCAL). We utilize the mediator variable based on Front-Door Criterion, to remove the confounding bias. Additionally, we adopt the pessimistic principle to tackle the distributional shift problem induced by the under-coverage issue. This issue refers to the mismatch of distributions between the action distributions induced by candidate policies, and the policy that generates the observational data (known as the behavior policy). Our key observation is that, by incorporating auxiliary variables that mediate the effect of actions on system dynamics, it is sufficient to learn a lower bound of the mediator distribution function, instead of the Q-function, to partially mitigate the issue of distributional shift. This insight significantly simplifies our algorithm, by circumventing the challenging task of sequential uncertainty quantification for the estimated Q-function. Moreover, we provide theoretical guarantees for the algorithms we propose, and demonstrate their efficacy through simulations, as well as real-world experiments utilizing offline datasets from a leading ride-hailing platform.</p><p dir="ltr">In the second part of this dissertation, in contrast to the first part, which approaches the distributional shift issue implicitly by penalizing the value function as a whole, we explicitly constrain the learned policy to not deviate significantly from the behavior policy, while still enabling flexible adjustment of the degree of constraints. Building upon the offline reinforcement learning algorithm, TD3+BC \cite{fujimoto2021minimalist}, we propose a model-free actor-critic algorithm with an adjustable behavior cloning (BC) term. We employ an ensemble of networks to quantify the uncertainty of the estimated value function, thus addressing the issue of overestimation. Moreover, we introduce a method that is both convenient and intuitively simple for controlling the degree of BC, through a Bernoulli random variable based on the user-specified confidence level for different offline datasets. Our proposed algorithm, named Ensemble-based Actor Critic with Adaptive Behavior Cloning (EABC), is straightforward to implement, exhibits low variance, and achieves strong performance across all D4RL benchmarks.</p>

The purpose of the present study was to investigate the impact of Same-Language Subtitles (SLS) on the subject-specific comprehension and the academic literacy levels of EAL students on the Vaal Triangle Campus of North-West University (NWU). Essentially, the study aimed to determine whether exposing students studying through English as an Additional Language (EAL) to subtitled lectures (live or recorded lectures) would help improve their comprehension of the academic content as well as their receptive academic literacy skills, compared to students who were not exposed to subtitled lectures. This study stems from the identification of an academic performance-related issue on the Vaal Triangle Campus of NWU. Indeed, campus statistics show that the throughput rate of EAL students remains low, and that these students‟ academic literacy levels are inadequate. In other words, EAL students on this campus are underachieving and seem to have difficulties in mastering academic English. Based on various studies that showed SLS to be a valuable tool in terms of learning and academic literacy, this study proposed to introduce SLS (both live via respeaking and offline) in the university classroom as a learning aid, thus optimising the time students spend in lectures. Very little information was available in South Africa on the impact of SLS on the subject-specific comprehension of EAL students in a tertiary academic context. Furthermore, it had been anticipated that a certain number of technical constraints were likely to be encountered during the empirical investigation. These two factors made it difficult to predict what other factors could influence the outcome of the study. As a result, the study was based on the principle of Action Research, a research method characterised by the fact that the research is carried out in as many cycles as may be necessary in order to achieve the optimal conditions for a specific intervention. Three cycles were necessary to reach the optimal design of the present study so that a confident conclusion could be made regarding the impact of SLS on comprehension and academic literacy. For each cycle, the intervention was carried out over an academic semester. In the first cycle, a test group composed of EAL first-year Economics students was exposed to live SLS via respeaking during class, while a control group (also composed of EAL first-year Economics students) attended the same class at a different time, without SLS. In the second cycle, the live SLS via respeaking were replaced with offline SLS. The intervention was taken out of the regular classes and was carried out in the context of practical revision classes scheduled specifically for the purpose of the intervention. The test group viewed subtitled videos of lectures, while the control group viewed videos without subtitles. After each viewing, all participants were required to complete a short comprehension test. This cycle was also conducted in first-year Economics. The basic design of the third cycle was similar to that of the second cycle, but for the fact that the intervention took place in the context of a Psychology module, which, unlike the Economics module, was taught without the lecturer making use of slides. After these three research cycles were completed, it could be concluded that offline SLS indeed have a positive impact on the subject-specific comprehension and the receptive academic literacy skills of EAL students in a tertiary academic context. This conclusion was supported by the following findings: 1. The first research cycle pointed towards a slight, but statistically insignificant benefit in terms of both comprehension and academic literacy. However, at this stage of the empirical investigation, the technical constraints made it difficult to draw a precise conclusion in that respect. 2. In the second research cycle, the SLS seem to have had a significant impact on the receptive academic literacy skills of the test group, compared to the control group. However, no such impact could be noted in terms of subject-specific comprehension. This was attributed to the presence of a confounding variable, namely slides used during the lectures. This once more made it impossible to draw a confident conclusion regarding the impact of SLS on comprehension. 3. The third research cycle made a more confident conclusion regarding the impact of SLS on subject-specific comprehension possible. Indeed, the results of the statistical analyses show that the test group performed significantly better in their semester test (covering the work done in all the recorded lectures) than the control group, which was not exposed to any videos at all. On the basis of these findings, it was concluded that SLS in their offline form have a positive impact on the subject-specific comprehension and the receptive academic literacy skills of EAL students in a tertiary academic context, specifically if the students are given sufficient time to get used to the mode. This study seems to indicate that the benefits of SLS for comprehension can be recorded provided that students are exposed to the intervention over a longer period of time. However, there may be further scope for refinement as far as this study is concerned. It is therefore important that the topic be investigated further. / MA, Language Practice, North-West University, Vaal Triangle Campus, 2012

在行動寬頻、雲端及網際網路等產業的帶動下，全球行動裝置如智慧型手機及平板電腦蓬勃發展，其衍生之行動經濟及應用更為全球經濟成長帶來高度動能。在連網裝置逐年成長之下，物聯網產業應運而生，而其中感測器技術更為物聯網基礎建設中最為核心的關鍵項目。繼蘋果於西元2013年WWDC大會發佈iBeacon技術後，全球零售業掀起微定位熱潮，而Forbes更預估企業利用Beacon搭配APP蒐集消費者資訊為未來物聯網重要趨勢之一。在實體零售店逐漸「展示店化」趨勢下，如何藉由ICT科技、物聯網技術、行動裝置應用來加強與消費者間的連結及提升購物體驗，為實體零售業者急需面對的課題。 本論文研究主要的目的在於探討Beacon於台灣零售業者燦坤的創新擴散歷程，以及使Beacon能夠於燦坤導入每階段快速擴散的關鍵因素，而其中以創新擴散模型之認知、說服、決策、實行、確認五階段構面進行研究與分析。此外，更探討燦坤如何運用Beacon進行O2O虛實整合。 本研究所得到的主要結論包括：(1)在創新擴散的流程中，擁有與導入科技相關的組織團隊背景以及根據創新需求調整組織架構，為Beacon於燦坤之認知階段中能夠快速擴散的關鍵因素。(2)在創新擴散的流程中，鼓勵創新的企業文化為Beacon於燦坤之說服階段中能夠快速擴散的關鍵因素。(3)在創新擴散的流程中，將非核心能力專案外包以及對於外包廠商的選擇，為Beacon於燦坤之決策階段中能夠快速擴散的關鍵因素。(4)在創新擴散的流程中，高階主管對於專案的參與支持以及選擇場域面積、營業額、人流數較大的門市進行首波產品曝光測試，為Beacon於燦坤之實行階段中能夠快速擴散的關鍵因素。(5)在創新擴散的流程中，進行產品的成效評估和顧客滿意度調查，以及尋求產品穩定的獲利模式來源，為Beacon於燦坤之確認階段中能夠快速擴散的關鍵因素。(6)燦坤以Beacon為技術核心打造燦坤黃金傳說APP，以門市尋寶及APP推播的方式，連結燦坤3C實體門市及快3網路商城，使虛實能夠結合及互利。本文最後並提出對於實務上及後續研究的建議。

Dagens teknologi tillåter inte att göra kopplingar i konsumentdata mellan hemsida och den  fysiska butiken. Det innebär att företag går miste om värdefull kunskap om sina kunders beteende och effekten av deras marknadsföringsåtgärder. Syftet med uppsatsen är att undersöka vilket hemsidebeteende som föregår köp i fysisk butik och hur beteendet skiljer sig jämfört med köp online. Uppsatsen syftar även till att undersöka hur hemsidebeteendet påv​erkar fysiska butikers ​konsumenters köpstorlek samt hur ​konsumenters köpsituation påverkar hemsidebeteendet. Studien gjordes i samarbete med ett möbel- och heminredningsföretag, och datainsamling skedde genom en enkätundersökning i butik, Google Analytics samt observationsstudie i samband med enkätundersökningen. Resultaten visar påtagliga skillnader i hemsidebeteende beroende på köpkanal och att konsumenter spenderar mer i fysisk butik om hemsidan besökts inför köp. Studien visar även att det är mer troligt att otåliga och engagerade konsumenter besöker hemsidan inför köp jämfört med konsumenter som värderar shoppingupplevelsen i sig.

Lesbian, gay, bisexual, transgender and intersex people (LGBTI) are discriminated and subjected to violence in societies across the world, and ensuring their rights is on the international agenda. On a European level, nationalism, state-led persecution and rhetoric of hate have slowed down the process of ensuring human rights for LGBTI people particularly in Eastern Europe, where they are subject to violence, discrimination and hostility. One type of hostility targeting LGBTI people is hate speech. As literature has shown, hate speech can have harmful effects on its targets. Additionally, with the increased accessibility and use of the Internet and social media networks, hate speech has become more widespread and new challenges have appeared. Our research objective has addressed the gap identified by Brown (2018); to contribute to the research on hate speech and its harmful effects in general, comparing the effects of online and offline hate speech in particular. Additionally, by comparing the findings from Moldova and Ukraine, we have investigated if the context in which the hate speech is produced has any effect on the harmful effects experienced by targets. The research was conducted using a mixed method with a parallel convergent design, giving equal priority to qualitative and quantitative data. Data collection took place in Moldova (Chisinau) and in Ukraine (Kyiv) during Pride in the respective countries. Due to the nature of this research, results are not representative, and conclusions drawn can neither be applied to the entire LGBTI community in Moldova nor in Ukraine. Conclusions can, however, provide interesting insights for further research. Constitutive and consequential harms from online and offline hate speech are experienced by the LGBTI activists and community in both countries. In terms of constitutive harms, LGBTI community have suffered from harms like negative impact on self-esteem, silencing, psychological distress and restrictions on freedom of movement and association. Experienced consequential harms were negative stereotyping, physical violence and normalization of discrimination. Further, this research indicated that there is a difference in terms of harms caused by online and offline hate speech when it comes to the constitutive harms, as the harms from offline hate speech seemed to be experienced to a larger extent. Comparing results from Moldova and Ukraine, it can be concluded that the content of hate speech and harmful effects of hate speech are experienced differently depending on the context in which hate speech was produced. In general, hate speech in Ukraine seemed to be more violent and aggressive while in Moldova it was more related to the structural violence.

繼網路創業後，近年來可以看到以行動應用創業的案例越來越多，而之中又以整合虛實通路的O2O (Online to Offline，線上至線下的通路模式)模式更為熱門。過去由於虛擬通路的盛行，造就了電子商務的快速發展，也因而使得消費資訊傳遞得更加快速。在這樣的情況下，企業的策略與經營模式也不斷地調整，電子商務方面則產生了很大的變化，原本只著重於經營虛擬通路的企業，開始反向經營實體通路，而原本只經營實體通路的企業，則開始經營起虛擬通路。 本論文研究主要的目的在於探討台灣新創公司如何經營O2O的模式，本研究以核心資源、營運範疇、事業網路等三構面，探討三家新創公司（食我餐飲顧問公司、百迪爾資訊公司、呼叫師傅公司）O2O方面的經營策略，並藉由第四構面-價值創造，探究該等公司在O2O服務上所創造的價值。 本研究所得到的主要結論包括：(1)新創公司在O2O的初期經營上，會著重於需求端。(2)新創公司在經營O2O的業務推廣上，會尋求外部資源的協助。(3)新創公司在經營O2O的服務品質控管上，會透過廠商篩選機制的建立以及與廠商的合約來加以強化。(4)新創公司在O2O的經營上，會透過在地資訊的即時提供來提高使用者採用O2O平台的意願。本文最後並提出實務上的意涵與後續研究的建議。 / Since the trend of running businesses on the Internet started to foster, we could see that there are more and more cases taking portable devices as the core of their business. Besides, the Online to Offline (O2O) business model of using the clicks-and-mortar channel might be one of the most popular choices. In the past, due to the prevailing of virtual channel, it did make e-commerce to be in vogue, and also help consuming information to transact more rapidly. In this kind of situations, the business strategies and business models of enterprises also kept on making adjustments to follow the flow. As for e-commerce, it did resulted in huge changes. For those enterprises who used to concentrate on virtual channels, they started the business of physical channels; those enterprises who only aiming at running physical channels started to explore the area of virtual channels. The main purpose of the study is to discuss about how those Taiwanese start-up companies operate O2O business. Moreover, the study takes core resources, operating area and business networks to view the O2O business strategies of three start-up companies. Also, by using value creation, the study wants to understand the value created by the O2O business of those companies. After the research of the issues above, the study came up with four main conclusions: 1. Start-up companies would concentrate on the demand side in the beginning of their business. 2. Start-up companies would seek for the help of outer resources when they were promoting their business. 3. In the aspect of controlling the quality of their services, start-up companies would choose to build the filtering system of target companies and signing contracts. 4. In running O2O business, start-up companies would offer local information immediately to raise consumers’ will of using their O2O platform. With those four main point provided by the study, it also brings up the meaning the practical meaning of the study and suggestions of following researches. Keywords: Start-up Company, O2O (Online to Offline), Business Strategy, Value Creation

Approximate string matching consists in identifying strings as similar even ifthere is a number of mismatch between them. This technique is one of thesolutions to reduce the exact matching strictness in data comparison. In manycases it is useful to identify stream variation (e.g. audio) or word declension (e.g.prefix, suffix, plural). Approximate string matching can be used to score terms in InformationRetrieval (IR) systems. The benefit is to return results even if query terms doesnot exactly match indexed terms. However, as approximate string matchingalgorithms only consider characters (nor context neither meaning), there is noguarantee that additional matches are relevant matches. This paper presents the effects of some approximate string matchingalgorithms on search results in IR systems. An experimental research design hasbeen conducting to evaluate such effects from two perspectives. First, resultrelevance is analysed with precision and recall. Second, performance is measuredthanks to the execution time required to compute matches. Six approximate string matching algorithms are studied. Levenshtein andDamerau-Levenshtein computes edit distance between two terms. Soundex andMetaphone index terms based on their pronunciation. Jaccard similarity calculatesthe overlap coefficient between two strings. Tests are performed through IR scenarios regarding to different context,information need and search query designed to query on a technicaldocumentation related to software development (man pages from Ubuntu). Apurposive sample is selected to assess document relevance to IR scenarios andcompute IR metrics (precision, recall, F-Measure). Experiments reveal that all tested approximate matching methods increaserecall on average, but, except Metaphone, they also decrease precision. Soundexand Jaccard Similarity are not advised because they fail on too many IR scenarios.Highest recall is obtained by edit distance algorithms that are also the most timeconsuming. Because Levenshtein-Damerau has no significant improvementcompared to Levenshtein but costs much more time, the last one is recommendedfor use with a specialised documentation. Finally some other related recommendations are given to practitioners toimplement IR systems on technical documentation.

The purpose of the present study was to investigate the impact of Same-Language Subtitles (SLS) on the subject-specific comprehension and the academic literacy levels of EAL students on the Vaal Triangle Campus of North-West University (NWU). Essentially, the study aimed to determine whether exposing students studying through English as an Additional Language (EAL) to subtitled lectures (live or recorded lectures) would help improve their comprehension of the academic content as well as their receptive academic literacy skills, compared to students who were not exposed to subtitled lectures. This study stems from the identification of an academic performance-related issue on the Vaal Triangle Campus of NWU. Indeed, campus statistics show that the throughput rate of EAL students remains low, and that these students‟ academic literacy levels are inadequate. In other words, EAL students on this campus are underachieving and seem to have difficulties in mastering academic English. Based on various studies that showed SLS to be a valuable tool in terms of learning and academic literacy, this study proposed to introduce SLS (both live via respeaking and offline) in the university classroom as a learning aid, thus optimising the time students spend in lectures. Very little information was available in South Africa on the impact of SLS on the subject-specific comprehension of EAL students in a tertiary academic context. Furthermore, it had been anticipated that a certain number of technical constraints were likely to be encountered during the empirical investigation. These two factors made it difficult to predict what other factors could influence the outcome of the study. As a result, the study was based on the principle of Action Research, a research method characterised by the fact that the research is carried out in as many cycles as may be necessary in order to achieve the optimal conditions for a specific intervention. Three cycles were necessary to reach the optimal design of the present study so that a confident conclusion could be made regarding the impact of SLS on comprehension and academic literacy. For each cycle, the intervention was carried out over an academic semester. In the first cycle, a test group composed of EAL first-year Economics students was exposed to live SLS via respeaking during class, while a control group (also composed of EAL first-year Economics students) attended the same class at a different time, without SLS. In the second cycle, the live SLS via respeaking were replaced with offline SLS. The intervention was taken out of the regular classes and was carried out in the context of practical revision classes scheduled specifically for the purpose of the intervention. The test group viewed subtitled videos of lectures, while the control group viewed videos without subtitles. After each viewing, all participants were required to complete a short comprehension test. This cycle was also conducted in first-year Economics. The basic design of the third cycle was similar to that of the second cycle, but for the fact that the intervention took place in the context of a Psychology module, which, unlike the Economics module, was taught without the lecturer making use of slides. After these three research cycles were completed, it could be concluded that offline SLS indeed have a positive impact on the subject-specific comprehension and the receptive academic literacy skills of EAL students in a tertiary academic context. This conclusion was supported by the following findings: 1. The first research cycle pointed towards a slight, but statistically insignificant benefit in terms of both comprehension and academic literacy. However, at this stage of the empirical investigation, the technical constraints made it difficult to draw a precise conclusion in that respect. 2. In the second research cycle, the SLS seem to have had a significant impact on the receptive academic literacy skills of the test group, compared to the control group. However, no such impact could be noted in terms of subject-specific comprehension. This was attributed to the presence of a confounding variable, namely slides used during the lectures. This once more made it impossible to draw a confident conclusion regarding the impact of SLS on comprehension. 3. The third research cycle made a more confident conclusion regarding the impact of SLS on subject-specific comprehension possible. Indeed, the results of the statistical analyses show that the test group performed significantly better in their semester test (covering the work done in all the recorded lectures) than the control group, which was not exposed to any videos at all. On the basis of these findings, it was concluded that SLS in their offline form have a positive impact on the subject-specific comprehension and the receptive academic literacy skills of EAL students in a tertiary academic context, specifically if the students are given sufficient time to get used to the mode. This study seems to indicate that the benefits of SLS for comprehension can be recorded provided that students are exposed to the intervention over a longer period of time. However, there may be further scope for refinement as far as this study is concerned. It is therefore important that the topic be investigated further. / MA, Language Practice, North-West University, Vaal Triangle Campus, 2012