Refine
Has Fulltext
- yes (59)
Year of publication
- 2021 (59) (remove)
Document Type
- Master's Thesis (36)
- Bachelor Thesis (13)
- Article (10)
Keywords
- COVID-19 pandemic (2)
- (Nicht-) Verstehen (1)
- Blind Hiring (1)
- Brier Skill Score (1)
- Bundestagswahl (1)
- Business Intelligence (1)
- Business Model (1)
- CDO pricing (1)
- CDU (1)
- CO2 (1)
Institute
- FB I - Wirtschaftswissenschaften (59) (remove)
Die Masterarbeit untersucht die Einflüsse personalisierter Werbung im Umfeld sozialer Medien auf die Wahrnehmung der Werbeanzeigen und das Online-Impulskaufverhalten deutscher Konsumenten.
Dazu wurde ein theoretisches Modell entwickelt, welches anhand einer Online-Befragung überprüft wurde. Die Auswertung der Ergebnisse erfolgte mittels konfirmatorischer Faktorenanalyse und eines Strukturgleichungsmodells.
The following thesis aims to explore the particular changes that occurred in user behavior on social media (Instagram, YouTube, Facebook) during the pandemic from March 2020 to present time, how these changes relate to user engagement rate and what impact these changes have on the online radio show industry from a perspective of emerging independent radio show series “Widows Radio”. For limitation purposes of the research, purposive sampling method is used to identify particular behaviors. As a result of the research, a list of changes of user behavior are identified, specific impact of these changes on user engagement rate is shown. On the base of the conducted research, several recommendations for the online radio company “Widows Radio” regarding their social media strategy are provided.
In the contemporary era, large number of companies publish their business performance as a report on the Internet. To make them usable in larger numbers, these reports must be read and labeled. This is inefficient and expensive. This study supports this business need by automatically classifying four different categories of metadata of financial reports. For training of the used random forest classifier, an active and a passive learning strategy are contrasted. The results show a clear advantage of active learning for classifying whether a financial report contains consolidated enterprise data or not. For more complicated multi-class classifications, no advantage is shown so far under the applied active learning strategy, however, there is potential to develop an alternative more efficient strategy.
Despite backlash and public outrage in recent years, the long existing phenomenon of crunch time as a method to speed up the progress of a video game development project persists in the video game industry. Regardless of its reputation, employees of the video game industry
continue to work under the use of this industry practice. The previous research has examined the reasons for the use of crunch time, but there is much to be explored concerning the point of view of the crunching employees as individuals with agency. Examining their perception presents an opportunity to better understand crunch as a multifaceted phenomenon. This knowledge can then be used to develop alternatives to using crunch time. Therefore the objective of this thesis is to determine the positive and negative impacts of crunch as well as potential alternatives from the employees’ perspective.
This thesis’ research uses semi-structured expert interviews with six expert gameworkers. The interviews are evaluated through the method of qualitative structured content analysis. It becomes apparent how the negative and positive perceptions are closely intertwined and
depend on the differing contexts and experiences of crunch time. In the case of the interviewed experts it was found that in some circumstances, an improved motivation, mental state and team spirit was perceived. In particular when employing agile methodologies, an
increased quality and productivity became apparent. In general the ability to meet deadlines was seen as positive. Negative perceptions were more commonly observed, such as damage to the team dynamics and spirit, negative impacts on the employees’ physical and mental
health, social life, work life balance, and creative ability. Furthermore, the long term mindset shift normalizing crunch time was noticed. Other impacts include a negative influence on product quality and scheduling, as well as brain drain due to skilled gameworkers leaving the studio or industry. Suggested alternatives target the conscious adjustment of the projects’ resources, schedule, and scope, as well as increased resources and improved management thereof. Furthermore the employees see the need for clear quality standards, and for managers to be more aware of their importance as role models. Other suggestions include the use of workshops and agile methodologies.
Purpose: This study aims at investigating how nonprofit organisations can successfully and sustainably implement web analytics. It develops a framework based on existing literature as well as expert interviews and tests its applicability and practicability in a hands-on case study with the German nonprofit organisation Gemeinsam TECHO e.V..
Methodology: The research questions are explored through a qualitative approach via a literature review and semi-structured interviews. The author draws a convenience sample of seven German nonprofit organisations. The interviews are analysed through qualitative content analysis. Based on the results a first web analytics framework is developed which is then tested and reviewed in a practical application.
Findings: The outcome of the study is a framework with a general step-by-step plan for NPOs on how they can implement web analytics into their business.
Research limitations: The convenience sample as well as the focus on German nonprofit organisations are possible limitations. The framework should be further tested with nonprofit organisations in different settings and conditions: country, industry, etc. Another limitation is that during the practical application only a single pass of the framework was done. Additional testing of the framework with several iterations might uncover more suggestions for improvement.
Value: This paper seeks to propose guidelines that give valuable hints for nonprofit organisations looking to implement web analytics into their business. Some existing web analytics frameworks can be found in current scientific literature; however, little empirical evidence exists for the case of nonprofit organisations. This study provides a new viewpoint on the use of web analytics which may be of use for those working in the context of nonprofit organisations.
Keywords: web analytics, nonprofit organisations, NPOs, TECHO, marketing, data analytics, web analytics framework
The use of the Country-of-Origin (COO) effect is a relevant topic in the field of international marketing research and is gaining more attention in today’s global economy. Especially in the cosmetics industry, the origin of brands tends to be included within their marketing strategy. Therefore, this thesis examines how the COO effect is impacting the evaluation of international cosmetics brands with influential market players such as the USA, South Korea and France. Using a quantitative approach, the opinions of consumers in relation to the different COO dimensions and the importance of these, as well as possible influencing factors and the impact on behavior, was analyzed. Ultimately, significantly different ratings of each dimension and their importance were found. Similarly, the influencing factors involvement and familiarity might affect the evaluation and it is likely that the higher the COO effect evaluation, the higher is the purchase intention and the willingness to pay. The findings underline the role of the COO effect for internationally marketed cosmetics and theoretical and managerial implications can be drawn based on these.
We are living in the age of data, organizations are facing exponential growth in data. This growth
of data if not dealt with in the right approach will do more harm than good to the organization.
One of the most important aspects of data is data quality. Increased data quality ensures better
decision-making, thereby enabling companies to stay competitive in the market. Data quality
directly affects business decision-making, consequently having a direct effect on the KPIs of the
organization. So it is integral for growing organizations to gain an understanding of where they
stand with regards to their data quality to make improvements needed so data quality problems
won’t affect the organization's goals and KPIs. The topic of data quality is a topic that has been
addressed in numerous researches and from different aspects, there are many established and
interesting studies that have been done and resulted in different frameworks that are followed in
our days now and that shows how rich and important this topic is. The thesis aims to gain a more
precise understanding of how data quality can affect the ability of an organization to achieve the
set key performance indicators (KPIs). To do that this study will take a live example which is the
company of Uberall and do a case study on it, the situation of data quality of Uberall was
analyzed through interviews with different employees from different departments. This research
gives an insight into some precise data quality problems that Uberall is currently facing and
needs to address .
The aim of this paper is to understand the current state of financial development in Mexico. First, the paper introduces the Financial Development Index (FDI) which provides an overview of the situation throughout the world, and then the paper analyzes several proxies for each aspect of financial development, namely access, depth, efficiency, and stability.
This thesis proposes a framework for extracting crucial textual information from the U.S. Securities and Exchange Commission annual 10-K reports using multiple natural language processing techniques. Ongoing progress has been made on retrieving the summary or sentiment from financial text in recent years due to the large volume of data and advanced computing power. However, the combination of different techniques for generating insight from 10-K textual content has not yet been addressed in the literature. Therefore, the main study contribution is designing a framework that integrates text summarization, sentiment analysis, topic modeling, and regression to provide readers with several indicators to help them acknowledge the valuable information in 10-K reports more efficiently and effectively. The framework considers various models but focuses on transformer-based models because the literature review indicates that these models surpass the performance of others on most natural language processing tasks. As a result, the framework successfully condenses the report content by 88% into the most critical sentences with the sentiment and discussed topic. Finally, these extracted features are further used to predict future company growth that, in addition to the traditional quantitative metrics, can be a new reference for investors in making more comprehensive decisions.
Abstract
Der Einsatz gezielter Content-Strategien im Kontext politischer Kommunikation auf Kam- pagnenebene kann wortwörtlich in letzter Sekunde Wahlen entscheiden. Nicht ohne Grund hat der Teilbereich der öffentlich wirksamen politischen Kommunikation in den letzten Jah- ren enorm an Bedeutung gewonnen. Dabei verschiebt sich nicht nur die politische Kommu- nikation derweilen mit enormer Geschwindigkeit ins Internet, sondern ganze Wahlkämpfe werden online ausgetragen. Insbesondere Social Media-Kanäle spielen eine bedeutende Rolle.
Um in einem informationsüberflutendem und von politischen Maximalpositionen übersät- tigtem Umfeld zu bestehen, wird es notwendig, herauszustechen und seine Zielgruppe tief- gehend zu verstehen, um sie auf dieser Basis durch zielgerichtete Inhalte anzusprechen und als Wähler (zurück) zu gewinnen. Eine vollumfängliche Kommunikationsstrategie endet je- doch nicht bei der Auswahl der Contents, sondern setzt darüber hinaus die richtigen Medi- enkanäle, Ansprechpartner und Art und Weise authentischer Kommunikation voraus. Konkreter Forschungsgegenstand der vorliegenden Masterarbeit ist das Gewinnen junger Menschen, der sogenannten Generationen Y und Z. Grundlage war dabei der aktuelle For- schungsstand hinsichtlich der Charakteristika dieser. Aus marketingpolitischer Sicht ist die- ses Forschungsfeld bereits sehr gesättigt, die Einzigartigkeit und somit Forschungslücke ent- stand durch das Überführen dieser Erkenntnisse auf politische, marketingadressierte Kom- munikation, mit Fokus auf die CDU Deutschlands. Auf dieser Basis wurde anhand durchge- führter quantitativer sowie qualitativer Forschung die Voraussetzungen erfolgreicher Kam- pagnenkommunikation geschaffen. Gegenstand und Ziel war dabei, eine Content-Strategie für eine Kampagne der CDU Deutschlands vor den Bundestagswahlen im September 2021 zu entwickeln, um die Attraktivität und Wählbarkeit für die Zielgruppe junger Menschen zu erhöhen. Es hat sich gezeigt, dass durch zielgerichtete Inhalte, authentische Vermittlung die- ser und die richtige Auswahl der Medienkanäle die angestrebte Zielerreichung realistisch erscheint. Weiterführende Forschungsfelder entstanden auf dieser Basis hauptsächlich im Bereich langfristiger politischer Loyalität durch politische Partizipation und die Erforschung hinsichtlich der Einstellung junger Menschen zu Parteien im Allgemeinen, auch und vor allem in Abhängigkeit demographischer Faktoren.
This thesis focuses on the automatic classification of research methods used in scientific articles in the domain of Information Systems, and studies the effectiveness of a state-of-the-art long-document Transformer technique, Longformer, on the multi-label classification task.
In recent years, the task of automatically extracting knowledge from academic articles has become more and more popular. However, as far as the author knows, due to the limitations of the max input sequence length of Transformer models, there has not yet been a comprehensive study on the classification of research methods using Transformer models, which have made extraordinary achievements in various Natural Language Processing (NLP) fields. Therefore, this thesis establishes an artifact that uses a modified Transformer model, Lomgformer, which can proceed with long-sequence inputs, to discuss its effectiveness and possible limitations. Additionally, this thesis also discusses and evaluates the performance of other benchmark models, such as the traditional Transformer - BERT and RoBERTa, and a non-transfer-learning model based on the architecture of convolutional neural network (CNN), which has been proven to have good performance by previous researchers.
As a result, this thesis proves that the Longformer-based artifact can effectively improve the classification performance for the scientific articles, and surpasses all the other models, not only the traditional Transformer models but also the models presented in the literature from the previous researchers.