Background: Developing new methods for exploring and analyzing data in virtual and augmented reality presents many opportunities and challenges, both in terms of software development and design inspiration. There are various hardware options, from Google Cardboard to Oculus Rift. Taking part in this challenge demands programming skills as well as creativity. A basic VR or AR application for exploring a specific type of open data will be developed by the student. The use of a platform-independent kit such as A-Frame is essential, as the application will be compared in a small user study to its non-VR version in order to identify advantages and disadvantages of the visualization method implemented. Details will be discussed with supervisor. Research problem: How can AR and VR be used to improve exploration of data? Some References: Butcher, Peter WS, and Panagiotis D. Ritsos. "Building Immersive Data Visualizations for the Web." Proceedings of International Conference on Cyberworlds (CW’17), Chester, UK. 2017. Teo, Theophilus, et al. "Data fragment: Virtual reality for viewing and querying large image sets." Virtual Reality (VR), 2017 IEEE. IEEE, 2017.

Millais, Patrick, Simon L. Jones, and Ryan Kelly. "Exploring Data in
Virtual Reality: Comparisons with 2D Data Visualizations." Extended
Abstracts of the 2018 CHI Conference on Human Factors in
Computing Systems. ACM, 2018.

More info at http://mitloehner.com/lehre/thesis-en.html

Recent years have seen an increasing availability of process execution data from several data sources. Process mining offers different analysis techniques to extract business insights from these data, known as event logs. An event log is a set of the executed process instances, i.e., cases. Each event in the event log represents an execution of process activity. The primary attributes associated with an event carry the control-flow information as execution timestamp, executed activity, and case identifier. Additionally, an event has different data attributes that describe business objects used for the process execution and process environment, e.g., resources, project name, location, cost..., etc.

The analysts use the process mining techniques to discover the control-flow model, identify the deviations using conformance checking techniques and provide recommendations and suggestions for process improvement.

Unfortunately, these techniques focus on the event process-attributes and ignoring the event data-context attributes. Analyzing the data-context attributes will improve the understanding of the process execution environment and the performed analysis quality.

The data-context attributes provide information about the business objects affected by executing the process instances. These attributes may be constant within a case or change over time in a case. Detecting the data-context attributes changing patterns will help understand what triggers these changes and how they affect the process variations. To detect the changing patterns, we need to identify the association and correlation rules between the data-context attributes. 

The thesis's objective is to build a tool that detects and visualizes the associate rules between the log's event data-context attributes.

References:

[1] Pentland, B., et al. "Bringing context inside process research with digital trace data." Journal of the Association for Information Systems 21 (2020).

[2] Pegoraro, Marco, Merih Seran Uysal, and Wil MP van der Aalst. "Discovering Process Models from Uncertain Event Data." International Conference on Business Process Management. Springer, Cham, 2019.

[3] Motahari-Nezhad, Hamid Reza, et al. "Event correlation for process discovery from web service interaction logs." The VLDB Journal 20.3 (2011): 417-444.

[4] Hornik, Kurt, Bettina Grün, and Michael Hahsler. "arules-A computational environment for mining association rules and frequent item sets." Journal of Statistical Software 14.15 (2005): 1-25.

Wikidata [1] has started to evolve from a side project of Wikipedia to a huge knowledge repository that allows collabvorative editing of data and it's schema by a large community of editors in a jointly developed Knowledge Graph.

The historical roots of Wikidata lie in similar "Semantic Wiki" Platforms such as Semantic Media Wiki [2], which as evolved into Wikidata's base underlying enginde calld Wikibase [3].

In this thesis we are interested in comparing tools like Semantic Media Wiki and Wikibase based on their functionality,, for instance in terms of creating data input forms, or importing existing schemata [4].

This theses shall resulting in a prototype for schema import or form construction for Wikibase.

[1] Vrandecic, D. Krötzsch,M. Wikidata: a freecollaborative knowledge base. Communications of the    ACM57(10), 78–85    (Sep 2014). https://doi.org/10.1145/2629489,https://dl.acm.org/doi/10.1145/2629489

[2] Krötzsch, M., Vrandecic, D., Völkel, M. Haller, H., Studer, R. 2006. Semantic Wikipedia. Proceedings of the 15th
International conference on World Wide Web, WWW 2006.
http://www.aifb.kit.edu/web/Inproceedings1055

[3] Diefenbach D., Wilde M.D., Alipio S. (2021) Wikibase as an Infrastructure for Knowledge Graphs: The EU Knowledge Graph. In: Hotho A. et al. (eds) The Semantic Web – ISWC 2021. ISWC 2021. Lecture Notes in Computer Science, vol 12922. Springer, Cham. https://doi.org/10.1007/978-3-030-88361-4_37

[4] https://github.com/TIBHannover/ontology2smw
 

Background:

The quality of the outcomes of process mining techniques strongly relies on the adequate input data. Domain knowledge plays an important role on the task of choosing and preparing the data. Usually process mining techniques rely on domain expert knowledge which can be error prone and not always available. Knowledge representation artifacts such as Ontology or Knowledge Graphs may support the process mining techniques

Research problem:

The core research problem addressed is: How process mining techniques can benefit from domain knowledge artifacts, such as ontologies or knowledge graphs?

This thesis can be conducted in three ways:

• With the aim of doing a systematic review of the literature on techniques that consider domain knowledge artifacts to support process mining.

• With the aim of proposing a method that uses knowledge graph for supporting process mining

• with the aim of doing a case study, i.e. applying process mining with the support of domain knowledge in a real-world case.

Requirements:

The candidate must have previous knowledge on process mining and knowledge graph. Further desirable requirements are pro-activity and self-organization. 

Initial references

1. Mohammad Khanbabaei, Farzad Movahedi Sobhani, Mahmood Alborzi, Reza Radfar: Developing an integrated framework for using data mining techniques and ontology concepts for process improvement. J. Syst. Softw. 137: 78-95 (2018).

2. Stefano Bistarelli, Tommaso Di Noia, Marina Mongiello, Francesco Nocera:
   PrOnto: an Ontology Driven Business Process Mining Tool. KES 2017: 306-315

 

Successful and innovation productions bring together the efforts of many
individuals. The patterns of their previous interactions, encoded in
their collaboration networks, are thought to be predictive of their
creativity and success. In this thesis, you will collect data on the
production history of the German-language television series Tatort. You
will map the collaboration networks of the key players (i.e. director,
writer, lead actors, cinematographer), and relate their structure to
outcomes, measured in terms of critical and commercial success. This
thesis will combine data collection (via webscraping), network science,
and regression modeling with theory from the social networks literature.

References
Uzzi, Brian, and Jarrett Spiro. "Collaboration and creativity: The small
world problem." American journal of sociology 111.2 (2005): 447-504.
Vedres, B., & Stark, D. (2010). Structural folds: Generative disruption
in overlapping groups. American Journal of Sociology, 115(4), 1150-1190.
De Vaan, M., Stark, D., & Vedres, B. (2015). Game changer: The topology
of creativity. American Journal of Sociology, 120(4), 1144-1194.
Lutter, Mark. "Do women suffer from network closure? The moderating
effect of social capital on gender inequality in a project-based labor
market, 1929 to 2010." American Sociological Review 80.2 (2015):
329-358.

The climate crisis is thought to both intensify natural disasters and to
extend their geographic coverage. A recent paper published in Nature
uses satellite data to estimate increases in flood exposure in recent
years, and to forecast future changes in flood risk. The goal of this
thesis is to apply the method and data specifically to Austria, creating
a detailed dataset of municipalities and regions that are at risk. This
dataset will be compared against out of sample data on the flooding that
occured in 2021.
References
Tellman, B., Sullivan, J.A., Kuhn, C. et al. Satellite imaging reveals
increased proportion of population exposed to floods. Nature 596, 80–86
(2021).

Online collaboration, especially in the software industry,increasingly takes place on a few centralized platforms.

GitHub and Slack are two distinguished examples. Occassionally these services
experience disruptions and are unavailable for hours or even days.
These shocks present an opportunity to study what developers do when
primary tools for collaboration and communication are offline. Do they
focus on alternative tasks? Do they explore new ideas? Or are times
 during outtages written off as lost? In this thesis, the student will
gather data on outtages of GitHub and Slack and compare activity on
before, during, and after these events. The student can study changes
in activity and behavior on GitHub and Stack Overflow, and in Google
searches of relevant terms using econometric methods.

Moldon, Lukas, Markus Strohmaier, and Johannes Wachs. "How
gamification affects software developers: Cautionary evidence from a
natural experiment on github." 2021 IEEE/ACM 43rd International
Conference on Software Engineering (ICSE). IEEE, 2021.
Malik, Momin M., and Jürgen Pfeffer. "Identifying platform effects in
social media data." Tenth International AAAI Conference on Web and
Social Media. 2016.
 Elmezouar, Mariam, et al. "Exploring the Use of Chatrooms by
 Developers: An Empirical Study on Slack and Gitter." IEEE Transactions
 on Software Engineering (2021).

Startups play an important role in the economy as innovators and new challengers to established firms. The goal of this thesis is to gather, clean, and analyze data from Crunchbase, a leading onlinesource for information about technology entrepreneurship. The focuswill be on the geographic distributions of startup activity and its relationship to other markers of innovation and technological development such as patents, economic complexity, and/or open source software activity.

 Yu, Sandy, and Lee Fleming. "Regional crowdfunding and high tech
entrepreneurship." Research Policy (2021): 104348.
Wachs, Johannes, et al. "The Geography of Open Source Software:
Evidence from GitHub." arXiv preprint arXiv:2107.03200 (2021).

Background
Knowledge graphs (KGs) are nowadays the main structured data
representation model on the web, representing interconnected knowledge of different
domains. There are several methods to model a KG. For instance, they can be
extracted from semi-structured web data, like DBpedia, or edited collaboratively by a
community, like Wikidata. Since there is no perfect method and knowledge about the
world is constantly changing, regular updates in the KG are required.
Community-based KGs rely on the input of their users community, which can lead to
the insertion of statements that violate the KG restriction rules.
Goal of the thesis
The goal of this thesis is to perform an empirical study on the mistakes done
on Community-based KGs.
Requirements
It's interesting that the student has some previous knowledge and interest in
databases, and knowledge graphs. Further desirable requirements are pro-activity
and self-organization.
Initial references
● Shenoy, K., Ilievski, F., Garijo, D., Schwabe, D., & Szekely, P. (2021). A Study
of the Quality of Wikidata. arXiv preprint arXiv:2107.00156.
● Amaral, G., Piscopo, A., Kaffee, L. A., Rodrigues, O., & Simperl, E. (2021).
Assessing the quality of sources in Wikidata across languages: a hybrid
approach. arXiv preprint arXiv:2109.09405.
● Vrandečić, D. (2012, April). Wikidata: A new platform for collaborative data
collection. In Proceedings of the 21st international conference on world wide
web (pp. 1063-1064).

Deep Learning has had tremendous success in business and science (LeCun, 2015, Schmidhuber, 2015, Bachhofner, 2020). The paper by (Krizhevsky, 2012) is widely accepted to be the key paper for this success, as they were able to substantially decrease the error on ImageNet Large Scale Visual Recognition Challenge 2012 (ILSVRC-2012), an image classification task, by using a convolutional neural network. Since this work, many scholars worked on scaling neural networks based on the idea that larger models lead to a better performance. One such scaling idea is the compound coefficient proposed by (Tan, 2019), which uniformly scales depth, width, and resolution. 

In this thesis, the student contributes to the question whether the compound coefficient has an influence on error patterns. In particular, to answer this question the students task is to perform a series of reproducible image classification experiments. 

Resources 

STAT 157, Introduction to Deep Learning, UC Berkley, https://courses.d2l.ai/berkeley-stat-157/index.html 

6.S191, Introduction to Deep Learning, Massachusetts Institute of Technology, introtodeeplearning.com/ 

Dive into Deep Learning, d2l.ai/ 

CS231n, Convolutional Neural Networks for Visual Recognition, Stanford University, http://cs231n.stanford.edu/ 

A collection of lecture on deep learning, Massachusetts Institute of Technology, deeplearning.mit.edu/ 

References 

LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436-444. 

Schmidhuber, J. (2015). Deep learning in neural networks: An overview. Neural networks, 61, 85-117. 

Bachhofner, S., Loghin, A. M., Otepka, J., Pfeifer, N., Hornacek, M., Siposova, A., ... & Hochreiter, R. (2020). Generalized Sparse Convolutional Neural Networks for Semantic Segmentation of Point Clouds Derived from Tri-Stereo Satellite Imagery. Remote Sensing, 12(8), 1289. 

Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems, 25, 1097-1105. 

Bilal, A., Jourabloo, A., Ye, M., Liu, X., & Ren, L. (2017). Do convolutional neural networks learn class hierarchy?. IEEE transactions on visualization and computer graphics, 24(1), 152-162. 

Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., ... & Rabinovich, A. (2015). Going deeper with convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 1-9). 

Tan, M., & Le, Q. (2019, May). Efficientnet: Rethinking model scaling for convolutional neural networks. In International Conference on Machine Learning (pp. 6105-6114). PMLR.

Deep Learning has had tremendous success in business and science (LeCun, 2015, Schmidhuber, 2015, Bachhofner, 2020). The paper by (Krizhevsky, 2012) is widely accepted to be the key paper for this success, as they were able to substantially decrease the error on ImageNet Large Scale Visual Recognition Challenge 2012 (ILSVRC-2012), an image classification task, by using a convolutional neural network. One of the key ideas the authors use is to model neurons as Rectifier Linear Units (ReLU). Among other things, authors state that they observed a significant speed up due to this neuron model. (Bilal, 2017) observed that errors by GoogleLeNet (Szegedy, 2015), an Inception based architecture, to not follow a random process but suggest that it implicitly learns the class hierarchy present in the output classes (the last layer of the network). 

In this thesis, the student contributes to the question whether different models of a neuron lead to different error patterns. In particular, to answer this question the students task is to perform a series of reproducible image classification experiments. 

Resources 

STAT 157, Introduction to Deep Learning, UC Berkley, https://courses.d2l.ai/berkeley-stat-157/index.html 

6.S191, Introduction to Deep Learning, Massachusetts Institute of Technology, introtodeeplearning.com/ 

Dive into Deep Learning, d2l.ai/ 

CS231n, Convolutional Neural Networks for Visual Recognition, Stanford University, http://cs231n.stanford.edu/ 

A collection of lecture on deep learning, Massachusetts Institute of Technology, deeplearning.mit.edu/ 

References 

LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436-444. 

Schmidhuber, J. (2015). Deep learning in neural networks: An overview. Neural networks, 61, 85-117. 

Bachhofner, S., Loghin, A. M., Otepka, J., Pfeifer, N., Hornacek, M., Siposova, A., ... & Hochreiter, R. (2020). Generalized Sparse Convolutional Neural Networks for Semantic Segmentation of Point Clouds Derived from Tri-Stereo Satellite Imagery. Remote Sensing, 12(8), 1289. 

Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems, 25, 1097-1105. 

Bilal, A., Jourabloo, A., Ye, M., Liu, X., & Ren, L. (2017). Do convolutional neural networks learn class hierarchy?. IEEE transactions on visualization and computer graphics, 24(1), 152-162. 

Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., ... & Rabinovich, A. (2015). Going deeper with convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 1-9).

Knowledge graphs are means to model a domain of interest via relationships between objects, where the objects are the nodes and the relationships are the edges of a graph. This is done with the intention to be knowledge (Hogan, 2020), e.g., in medicine (Rotmensch, 2017). These nodes and edges can be further mapped into a continuous vector space – which is referred to as embedding a knowledge graph (Wang, 2017). Embeddings are, however, not limited to nodes and edges but are also done for substructures (subset of nodes and/or edges) and even the whole-graph (Cai, 2018). Manufacturing companies hence might find the knowledge graph properties, i.e., explicit modelling of relationships and ability to embed the graph, helpful in achieving business goals, e.g., when multiple sensors capture data on an object which implies dependent measurements (Garofalo, 2018). 

The objective of this thesis is to conduct a comprehensive literature review on applications of knowledge graph embeddings in the context of manufacturing and Industry 4.0. In the thesis, the student will systematically identify and characterize use cases upon multiple dimensions (e.g., industry type, type of process, etc.) which might reveal interesting patterns, and even new use cases. The student might also be asked to give her/his opinion on the current state of the art. The review shall include grey literature, i.e., non-scientific pieces of literature, e.g., company websites, marketing artefacts, and the likes. 

Supervisors: Stefan Bachhofner, Elmar Kiesling 

Resources 

CS 520, Knowledge Graphs, Stanford University, https://web.stanford.edu/class/cs520/, web.stanford.edu/~vinayc/kg/notes/Table_Of_Contents.html 

References 

Hogan, A., Blomqvist, E., Cochez, M., d'Amato, C., de Melo, G., Gutierrez, C., ... & Zimmermann, A. (2020). Knowledge graphs. arXiv preprint arXiv:2003.02320. 

Rotmensch, M., Halpern, Y., Tlimat, A., Horng, S., & Sontag, D. (2017). Learning a health knowledge graph from electronic medical records. Scientific reports, 7(1), 1-11. 

Wang, Q., Mao, Z., Wang, B., & Guo, L. (2017). Knowledge graph embedding: A survey of approaches and applications. IEEE Transactions on Knowledge and Data Engineering, 29(12), 2724-2743. 

Cai, H., Zheng, V. W., & Chang, K. C. C. (2018). A comprehensive survey of graph embedding: Problems, techniques, and applications. IEEE Transactions on Knowledge and Data Engineering, 30(9), 1616-1637. 

Garofalo, M., Pellegrino, M. A., Altabba, A., & Cochez, M. (2018). Leveraging knowledge graph embedding techniques for industry 4.0 use cases. arXiv preprint arXiv:1808.00434. 

Type of thesis: Master thesis

Feature integration theory states that color is perceived in the first stage of attention enabling us to recognize it in a fast and efficient manner. Additionally, paper by Moody (2009) suggests that color is perceived three times faster than shape and that discriminability of symbols within conceptual models plays an important role in their understanding. Despite this, the results of studies on the color’s impact on conceptual model understanding are contradicting. Furthermore, there is a lack of studies that investigated how different color schemes, and the different ways of coloring parts of model play a role in model understanding.

The goal of this project is therefore to conduct the experiment that will investigate the impact of color scheme and color paths on conceptual model comprehension.

For further questions, please contact Djordje Djurica (djordje.djurica@wu.ac.at)

Moody, D. (2009), “The physics of notations: Toward a scientific basis for constructing visual notations in software engineering”

Reijers, H.A., Freytag, T., Mendling, J. and Eckleder, A. (2011), “Syntax highlighting in business process models”, Decision Support Systems

Kummer, T. F., & Mendling, J. (2021). The Effect of Risk Representation Using Colors and Symbols in Business Process Models on Operational Risk Management Performance. Journal of the Association for Information Systems, 22(3), 7.

Ikujiro Nonaka is regularly regarded as the academic pioneer of knowledge management (Serenko & Bontis, 2004). His seminal papers and books have had a lasting impact on the academic field. However, his legacy has an impact far beyond the field of knowledge management.

In this bachelor thesis, you will trace the intellectual phases of Nonaka's work and analyze the impact of his seminal works on fields beyond of knowledge management. Using a bibliometric analysis (Ferreira, 2018), you will identify research fields that draw on his work. The results you seek will not only highlight adjacent areas to knowledge management, but also point to potential research fields that are receptive to further developments of Nonaka's thinking.

Serenko, A., & Bontis, N. (2004). Meta‐review of knowledge management and intellectual capital literature: Citation impact and research productivity rankings. Knowledge and process management, 11(3), 185-198.

Ferreira, F. A. (2018). Mapping the field of arts-based management: Bibliographic coupling and co-citation analyses. Journal of Business Research, 85, 348-357.

The concept of phronesis, dating back to Aristotle, has recently been „rediscovered“ and has entered the stage of knowledge management (Nonaka & Takeuchi, 2019, 2021). In essence, it is about doing the right thing in a particular context to promote the common good. However, the concept remains theoretically elusive and empirically difficult to test. In a recent attempt to address this shortcoming, Raysa et al. (2021a, 2021b, in prep.) propose the Organizational Phronesis Scale (OPS). In this bachelor thesis, you will theoretically related organizational phronesis to (another) concept(s) (e.g., organizational purpose) and are among the first to use the OPS in combination with another/other scale(s) you identify in the literature. You will perform basic statistical analysis (correlation) of empirical data, that you will need to obtain, and discuss your findings.

Nonaka, I., & Takeuchi, H. (2021). Humanizing strategy. Long Range Planning, 102070.

Nonaka, I., & Takeuchi, H. (2019). The wise company: How companies create continuous innovation. Oxford University Press.

Rocha G. R., Pinheiro, P., D‘Angelo, M., & Kragulj, F. (2021) Organizational Phronesis Scale Development. 22nd European Conference on Knowledge Management - ECKM 2021

Rocha G. R., Pinheiro, P., Kragulj, F., & Nunes C. (2021) There remains much to learn about organizational phronesis. Theory and Applications in the Knowledge Economy - TAKE 2021

Rocha G. R., Pinheiro, P., Kragulj, F., & Nunes C. (in prep.) ONE STEP TOWARDS RECOGNIZING THE PRACTICALLY WISE COMPANY: MEASUREMENT AND VALIDITY

Different people like different pieces of art. They are attracted by the particular aesthetic that, for example, a painting provides. The same holds true for many other things like buildings, cars, and other products but also for intangible things like music or social environments. People prefer what they find aesthetically appealing, thus there likely is an aesthetic to almost everything in the world. Accordingly, there are also aesthetics in the field of business and management. A recently published Review in the Academy of Management Annals journal takes up the topic of aesthetics in organizations and presents an overview of the current body of literature on the topic (Baldessarelli et al., 2021). There you can also find other references to other reviews on the topic of aesthetics in organizations. Your task in this thesis will be to analyze the papers presented in the AOM article and to conduct a Rapid Structured Literature Review according to Armitage and Keeble-Allen (Armitage and Keeble-Allen, 2008) with particular focus on the research methods used to set up inquiry into aesthetics. This means that you will have to pay close attention the method sections of the articles that you review in order to collect and summarize methods that can inform our knowledge on organizational aesthetics.

Armitage, A., Keeble-Allen, D., 2008. Undertaking a Structured Literature Review or Structuring a Literature Review: Tales from the Field 6, 12.

Baldessarelli, G., Stigliani, I., Elsbach, K., 2021. The Aesthetic Dimension of Organizing: A Review and Research Agenda. Acad. Manag. Ann.

Nowadays, business processes are an integral part of almost every organization. The companies are thereby supported by information systems that provide a means to manage and monitor the execution of the processes. To store and query process relevant data, these information systems are often relying on relational databases. However, in comparison to relational databases, newer graph-based databases might provide more native support to store process relevant data. This stems from the fact that graph-based databases natively support the storage and querying of graphs, which processes, in their simplest form, are.

The research problem of this thesis is to investigate how and to which extend graph-based databases can be used to store business process relevant data. For this, the literature has to be analyzed in the field of business processes and graph-based databases. Building on this literature research, an evaluation of different graph-based databases for the suitability of storing and querying of business process relevant data has to be done.

References

Esser, S., & Fahland, D. (2020). Multi-Dimensional Event Data in Graph Databases. arXiv preprint arXiv:2005.14552.

Waibel, P., Novak, C. Bala, S., Revoredo, K., and Mendling, J.. Analysis of Business Process Batching using Causal Event Models. In EdbA’20 workshop at ICPM 2020.

Hogan, A., Blomqvist, E., Cochez, M., d'Amato, C., de Melo, G., Gutierrez, C., et. al & Zimmermann, A. (2020). Knowledge graphs. arXiv preprint arXiv:2003.02320.

Kumar Kaliyar, R. (2015, May). Graph databases: A survey. In International Conference on Computing, Communication & Automation (pp. 785-790). IEEE.

"Data is the new oil", but also a higlhy volatile asset. E.g., the value of data may increas or decrease depending on how many people have access to it, how recently it has been produced, etc. [1]
In this thesis, we would like you to review survey and categorize "monetization" models for data and information, as well as methods to assess the value of data in different contexts and use cases.

1. D. Laney. Infonomics: How to Monetize, Manage, and Measure Information as an Asset for Competitive Advantage. 2017

Background
Knowledge graphs (KGs) are nowadays the main structured data
representation model on the web, representing interconnected knowledge of different
domains. There are several methods to model a KG. For instance, they can be
extracted from semi-structured web data, like DBpedia, or edited collaboratively by a
community, like Wikidata. Since there is no perfect method and knowledge about the
world is constantly changing, regular updates in the KGs are required. Classes
defined by the Wikidata KG can have millions of instances, however, in order for
abstractions to be helpful, the number of subclasses of a given class should be
relatively limited.
Goal of the thesis
The aim of this thesis is to investigate the viability of cluster-based techniques
in creating groups of subclasses that can be used to reduce the amount of direct
subclasses in Wikidata knowledge graph, creating a more concise and organized
class hierarchy.
Requirements
It's interesting that the student has some previous knowledge and interest in
databases, and knowledge graphs. Further desirable requirements are pro-activity
and self-organization.
Initial references
● Ritchie, A., Chen, J., Castro, L. J., Rebholz-Schuhmann, D., & Jimenez-Ruiz,
E. (2021, April). Ontology Clustering with OWL2Vec. In CEUR Workshop
Proceedings. CEUR Workshop Proceedings.
● Shenoy, K., Ilievski, F., Garijo, D., Schwabe, D., & Szekely, P. (2021). A Study
of the Quality of Wikidata. arXiv preprint arXiv:2107.00156.
● Vrandečić, D. (2012, April). Wikidata: A new platform for collaborative data
collection. In Proceedings of the 21st international conference on world wide
web (pp. 1063-1064).

Background: 

Semantic technologies and knowledge graphs have seen widespread adoption on the web, where they are used in applications such as contextual search, user profiling, social network modeling, and question-answering systems. Industry uptake in the manufacturing domain, while less pronounced, has also been gaining momentum recently as indicated by a growing number of workshops and contributions on the topic. This suggests that the potential of semantic representations to tackle the many and varied data integration, interoperability, and AI challenges that the manufacturing domain faces in the context of Industry 4.0 is increasingly being recognized.

Research problem:

Standardized schemas, conceptualizations, and vocabularies are a key building block in the process of creating (internal or public) industrial knowledge graphs, as they facilitate integration of data from various sources within the production environment. The goal of this thesis is to review the state-of-the art in semantic representation in Industry 4.0.

To this end, you will conduct research into common categories of manufacturing data and (i) identify candidate standards, schemas, and vocabularies to represent data within these categories, (ii) discuss their potential to facilitate exchange of manufacturing data and/or to break down manufacturing data silos, and (iii) identify important gaps in available vocabularies necessary to build manufacturing knowledge graphs. You will thereby organize the landscape of existing standards, compare and contrast their coverage, identify overlaps and gaps, and analyze strengths and weaknesses.

Initial references:

1. Hogan, A., Blomqvist, E., Cochez, M., D’Amato, C., Melo, G.D., Gutiérrez, C., Gayo, J.E., Kirrane, S., Neumaier, S., Polleres, A., Navigli, R., Ngomo, A.N., Rashid, S.M., Rula, A., Schmelzeisen, L., Sequeda, J., Staab, S., & Zimmermann, A. (2020). Knowledge Graphs. ArXiv, abs/2003.02320. https://arxiv.org/abs/2003.02320

2. Varish Mulwad and Raghava Mutharaju. 2017. Industrial Knowledge Graphs Workshop 2017: co-located with the 9th International ACM Web Science Conference 2017 (Preface; part of Content). In Proceedings of the 2017 ACM on Web Science Conference (WebSci '17). ACM, New York, NY, USA, 423-424. DOI: https://doi.org/10.1145/3091478.3162382

3. I. Grangel-González, L. Halilaj, S. Auer, S. Lohmann, C. Lange and D. Collarana, "An RDF-based approach for implementing industry 4.0 components with Administration Shells," 2016 IEEE 21st International Conference on Emerging Technologies and Factory Automation (ETFA), Berlin, 2016, pp. 1-8, doi: https://doi.org/10.1109/ETFA.2016.7733503.

4. P. Adolphs, H. Bedenbender, D. Dirzus, M. Ehlich, U. Epple, M. Hankel, et al., "Reference architecture model industrie 4.0 (rami4.0)", ZVEI and VDI Status Report, 2015. Available: https://www.zvei.org/fileadmin/user_upload/Presse_und_Medien/Publikationen/2016/januar/GMA_Status_Report__Reference_Archtitecture_Model_Industrie_4.0__RAMI_4.0_/GMA-Status-Report-RAMI-40-July-2015.pdf

 

Additional example references:

• N. Petersen, L. Halilaj, I. Grangel-González, S. Lohmann, C. Lange, and S. Auer. ”Realizing an RDF-based information model for a manufacturing company – A case study”. In: The Semantic Web – ISWC 2017, pages 350–366. Springer International Publishing, 2017. Available: https://link.springer.com/chapter/10.1007/978-3-319-68204-4_31

• Mörzinger, Benjamin, et al. "A large-scale framework for storage, access and analysis of time series data in the manufacturing domain." Procedia CIRP 67 (2018): 595-600.
  Available: https://www.sciencedirect.com/science/article/pii/S2212827117312131

• Annane, Amina, Nathalie Aussenac-Gilles, and Mouna Kamel. "BBO: BPMN 2.0 Based Ontology for Business Process Representation." 20th European Conference on Knowledge Management (ECKM 2019). 2019.

• Kalaycı, Elem Güzel, et al. "Semantic integration of Bosch manufacturing data using virtual knowledge graphs." International Semantic Web Conference. Springer, Cham, 2020.
  Available: http://www.ghxiao.org/publications/2020-iswc-bosch.pdf

• Lemaignan, Severin, et al. "MASON: A proposal for an ontology of the manufacturing domain." IEEE Workshop on Distributed Intelligent Systems: Collective Intelligence and Its Applications (DIS'06). IEEE, 2006.
  Available: https://academia.skadge.org/publis/Lemaignan2006.pdf [early work]

• Lin, Hsiao-Kang, and Jenny A. Harding. "A manufacturing system engineering ontology model on the semantic web for inter-enterprise collaboration." Computers in Industry 58.5 (2007): 428-437.
  Available: https://www.sciencedirect.com/science/article/pii/S0166361506001837  [early work]

Background:

Named-entity recognition (NER) – also known as named-entity identification/chunking/extraction – is a subtask of information extraction that seeks to locate and classify named entities mentioned in unstructured text into pre-defined categories such as person names, organizations, locations, medical codes, time expressions, quantities, monetary values, percentages, etc. (https://en.wikipedia.org/wiki/Named-entity_recognition) In particular, as NER is a task that is often preceded in the process of performing Natural language processing (NLP), it can be seen as the most primary task in constructing a knowledge graph and operating related artificial intelligence applications afterwards. Therefore, using more precise NER model while minimizing errors is significant for the effective operation of advanced computer applications to be added later.

Research problem:

The goal of this thesis is to conduct (1) a literature review and (2) organize the state-of-the-art in Named-entity recognition in computer sciences.

Initial references:

1. Nadeau, D., & Sekine, S. (2007). A survey of named entity recognition and classification. Lingvisticae Investigationes, 30(1), 3-26.

 

https://nlp.cs.nyu.edu/sekine/papers/li07.pdf

1. Sekine, S., Sudo, K., & Nobata, C. (2002, May). Extended named entity hierarchy. In LREC.

 

https://nlp.cs.nyu.edu/publication/papers/sekine-lrec02.pdf

1. Nothman, J., Ringland, N., Radford, W., Murphy, T., & Curran, J. R. (2013). Learning multilingual named entity recognition from Wikipedia. Artificial Intelligence, 194, 151-175.

 

https://www.sciencedirect.com/science/article/pii/S0004370212000276