330 Wirtschaft
Refine
Document Type
- Bachelor Thesis (3)
- Master's Thesis (2)
- Article (1)
Language
- English (6) (remove)
Is part of the Bibliography
- no (6)
Institute
Developing an intelligent chatbot that can imitate human-to-human interaction has become important in recent years. For this reason, many studies have been conducted to evaluate the quality of chatbots. Furthermore, various approaches and tools, such as sentiment analysis, have been created to improve the performance of chatbots.
This study examines previous research to identify the quality dimensions used to measure chatbots performance in order to develop a general chatbot assessment model that evaluates and compares chatbots quality. The developed evaluation model measures ten chatbot quality dimensions. This model is based on user experience. It requires human testers to interact with the chatbot to test its functioning and then a quantitative approach is used to collect data from user testing by conducting a survey with these testers. In this survey, they are instructed to evaluate the quality of the chatbot using a questionnaire that contains the items needed to evaluate each dimension.
This study also investigates whether sentiment analysis can improve the quality of chatbots and, if so, to identify the dimensions improved with sentiment analysis. For this reason, two chatbot versions are implemented using the Rasa framework (one that cannot detect sentiment and the other that analyzes sentiment and responds accordingly).
Following that, we used our evaluation model to evaluate and compare the two chatbot versions with two groups of participants by conducting a survey. In this survey, each group tested the functioning of one version. Then, both groups were instructed to use the items of the evaluation model to evaluate the version they tested. The goal of this survey was to evaluate the validity and reliability of the items used in the evaluation model to evaluate chatbots, and also to determine if sentiment analysis improved the chatbot quality by comparing survey results between the two groups. The results show that items used in the assessment model to evaluate chatbots are valid and reliable. The findings also indicate that sentiment analysis improves the chatbot’s quality. However, it improves the quality of some dimensions but not the majority of them.
In the course of this thesis, an overview will be given on which way developers can guide
users into acting environmentally friendly without the users realizing they are being
nudged. In the last couple of years, our private and work-life have been more and more
shifted away from reality into a digital context. Since the start of the Covid – 19 pandemic
in 2019, even more aspects of everyday life have been shifted to an online context, one
of them being groceries shopping. Even though online groceries shopping is not yet
common in Germany, there is a trend toward the online purchase of groceries visible.
This can be seen as a possibility to tackle another challenge the world is facing, the
climate crisis. One reason for the climate crisis is mindless consumption and purchasing
of too much food. This paper aims to combine the need for more aware consumption
with the newly rising trend of online supermarkets. Furthermore, a supermarket will be
provided to show if the implementation of environmentally–friendly nudges is technically
possible. To eventually prove the effectiveness of a nudge, it needs to be tested.
Keywords: Nudging, Environment, Online supermarkets
Artificial intelligence (AI) is one of the most auspicious yet controversial technologies with virtually unlimited potential to solve almost all of the existential problems humanity is facing today.1 Huge resources are poured into the development, testing and application of AI that is supposed to be utilized in almost all areas of everyday life.2 It may be used to combat genetically inherited diseases, to revolutionize the economy, to bring prosperity and equality to everyone and to counter the effects of climate change.3 With AI as the enabling technology humanity may experience a better future. Today, AI capabilities can already drastically improve analytic processing tasks and algorithmic systems and have beaten humans in games such as chess.4 Yet, AI and all of its applications bring about a myriad of ethical challenges. Revolutionary weapon systems that achieve autonomy via AI and genome-editing powered by AI are just some specific examples.5 An omnipotent AI will be either the greatest or the vilest thing that has happened to humanity in its brief existence.6 However, even today more and more computational devices are connected to each other, spurring a huge increase in global data streams that can be used to further train and enhance AI systems.
The prowess of AI for executing analytic tasks paves the way for the use of AI in more and more applications. One of these applications, that shows great promise, is the use of AI in surveillance applications.7 AI surveillance applications are proliferating at a fast rate, with a number of appli-cations already being in use today.8 These applications are aimed at accomplishing a number of policy objectives, some are in accordance with basic human laws, some are definitely not and some
1 Cf. Hawking (2018). P. 183ff
2 Cf. Hawking (2018). P. 183ff.
3 Cf. Hawking (2018). P. 183ff.
4 Cf. Burton (2015). P. 1ff.
5 Cf. Hawking (2018). P. 183ff.
6 Cf. Hawking (2018). P. 183ff.
7 Cf. Feldstein (2019). P. 1.
8 Cf. Feldstein (2019). P. 1.
2
belong in the nebulous area in between lawful and unlawful.9 But what are lawful and unlawful uses of AI surveillance systems and what are their ethical implications?
This thesis will examine the ethical implications of AI based mass surveillance systems and try to answer the first central question, if it is possible to use AI based mass surveillance applica-tions in an ethical way. Furthermore, the thesis will attempt to answer the second central ques-tion and find out how the ethical use of AI based mass surveillance systems, if this ethical use is possible, materialize. Governmental agencies will be in the focus of this discussion, as their use of the technology may have bigger ethical challenges. Yet private companies will play a part as well. In an attempt to accomplish these two aims, the thesis will inspect the basics of ethics and possible ethical theories that can be utilized to answer the questions. Normative ethics will be stud-ied first with a focus on consequentialism and utilitarianism. To gain a deeper understanding of utilitarianism, act and rule utilitarianism will be compared. Afterwards, deontological theories will be the focus of the discussion with a concentration on deontological pluralism. Next, the mentioned theories will be evaluated, discussing advantages and weak spots of the theories, to assess which theory may serve as the ethical framework of this thesis and the subsequent answer to the two main questions.
The next step will be the establishment of the AI framework. This contains the definition of AI and a distinction of terms that are commonly used in the its environment such as automation and au-tonomy. The importance of data for AI will be discussed. Afterwards, the technological basis of AI will be outlined, discussing key concepts such as machine learning and deep learning. Addi-tionally, it will be examined how an AI learns. The possible uses of AI in general will be outlined in a brief fashion, blazing the trail to discussing the moral challenges of AI. Afterwards, the current pace of AI development will be studied.
In the chapter that follows, the use of AI in surveillance technology is going to be highlighted. The possible ways of how AI can be used for surveillance purposes are reviewed here, discussing facial
9 Cf. Feldstein (2019). P. 1.
3
and behavioral recognition systems, smart cities, smart policing, communications/data driven sur-veillance and their enabling technologies. Then, the global proliferation of AI surveillance systems is going to be outlined.
Subsequently, the accordance of AI surveillance with basic human laws and rights, such as the right to privacy, will be checked to find out if the law and the international framework of human rights allow for AI surveillance or at least have restrictions that would greenlight the use of AI surveillance technology. All the aspects of the thesis, especially including the selected ethical framework, will be combined in this last section in order to enable the adaptation of a framework that allows to find out, if AI surveillance systems can be ethically permissible while also creating insights how this ethical AI surveillance system must be engineered. To finish, the thesis will end with a conclusion.
In this document a reliable data streaming mechanism for a TDMA LPWAN application is developed by adapting a link layer solution for power line communication, published at the International Symposium on Power Line Communications and its Applications (ISPLC) 2015. A C++ implementation of the services link layer is provided and demonstrated
working at a packet error rate of 50%.
This work aims to generate synthetic electromyographic (EMG) signals using Generative Adversarial Network (GAN). GANs are considered as one of the most exciting and promising approaches in deep learning [6], offering the possibility to generate artificial data based on real data. GAN consists of two main parts, a discriminator that attempts to differentiate between the generated data and the original data, and a generator that tries to fool the discriminator by generating data which looks like real data, the GAN works by staging a two-player
minimax game between generator and discriminator networks. To achieve the objective of generating realistic artificial electromyographic signals, two different architectures are considered for the generator and the discriminator networks of the GAN model: Long short-term memory (LSTM), which can avoid the long-term dependency problem and remembers information over a long period of time, and convolutional neural network (CNN), which is a powerful tool at automatic feature extraction. Different combinations of CNN and LSTM including hybrid model are experimented within the GAN using the same training data-set. The results and performances of each combination are compared and reviewed. The generated artificial EMG signals can be used to
simulate real muscle activity situations to for example improve muscle signal controlled prostheses using artificial data that may include conditions that does not exist in real data. This method of artificial data generation is not limited to EMG signals, the network can also be used to generate other synthetic biomedical signals such as electroencephalogram (EEG) or electrocardiogram (ECG) that can be practically used for testing algorithms and classifiers.