Categories
AI News

The award goes to…

13 December 2023

The award goes to....  

PhD Student Axel Christfort and Supervisor Associate Professor Tijs Slaats from the University of Copenhagen won the Process Discovery Contest at the 5th International Conference on Process Mining with their DisCoveR miner.

In a remarkable achievement, PhD student Axel Christfort and his supervisor, Associate Professor Tijs Slaats, won the Process Discovery Contest at the 5th International Conference on Process Mining.

Their cutting-edge DisCoveR miner produced the most accurate models and stood as the sole algorithm to successfully complete discovery and classification tasks within the stipulated time.

Process discovery algorithms play a crucial role in analyzing event logs, generating human-readable models that elucidate the behavior captured in the log. This includes understanding how individuals sequence activities in their work processes. The ICPM Conference, organizers of the Process Discovery Contest, evaluate submissions based on accuracy, requiring participants to mine models for a diverse range of logs and correctly classify corresponding ground truth traces.

This is the third prize in the Process Discovery Contest for the Process Modelling and Intelligence group from the Department of Computer Science, University of Copenhagen. In 2021, they secured awards for the best overall and the best imperative miner. The DisCoveR miner.

DisCoveR originated from a M.Sc. thesis by Viktorija Sali and Andrew Tristan Parli, supervised by Professor Slaats. The algorithm has undergone further refinement by Industrial PhD Student Christoffer Olling Back from ServiceNow, with ongoing enhancements by Axel Christfort. Funding from Independent Research Fund Denmark, DIREC – Digital Research Centre Denmark, and Innovation Fund Denmark has been instrumental in supporting this groundbreaking work.

Axel Christfort and Tijs Slaats are nominated Process Discovery Contest Winners

The industrial application of DisCoveR has been demonstrated through its implementation by DCR Solutions. The algorithm’s efficacy and utility have been validated in real-world scenarios, emphasizing its practical significance. Ongoing contributions from PhD Vlad Paul Cosma and Professor Thomas Hildebrandt have further extended and improved the miner, adding to its robustness.

Looking ahead, the Process Modelling and Intelligence group is eager to build upon these achievements to secure additional funding and foster novel collaborations. The team is already gearing up for the next iteration of ICPM, aiming to continue their winning streak and further advance the field of process discovery.

FACTS

Associate Professor Tijs Slaats is the project manager of the DIREC project ‘AI and Blockchain for complex business processes’.

Together with industry, the project aims to develop methods and tools that enable the industry to develop new efficient solutions for exploiting the huge amount of business data generated by enterprise and blockchain systems, with a specific focus on tools and responsible methods for the use of process insights for business intelligence and transformation.  

Categories
Business innovation Digital entrepreneurship News

It requires collaboration with people of different expertise to push your idea forward

11 December 2023

It requires collaboration with people of different expertise to push your idea forward  

Kurt Nielsen is one of the pioneers behind encryption and blockchain technology for protecting sensitive data while in use, and co-founder of Partisia. Here, he tells about the journey from researcher to CEO.

It all began as two research projects at Aarhus University, where a small team of researchers with diverse backgrounds in cryptography, business economics, and software development joined forces. The collaboration resulted in a groundbreaking cryptography technology, and in 2008, the tech company Partisia was born.

The CEO and partner Kurt Nielsen with a background in mathematics and economics, was involved from the beginning – from the first idea conceived by Professor Ivan Damgård at the Department of Computer Science at Aarhus University to the foundation of Partisia, which is now a leader in advanced cryptography and blockchain technologies for the financial sector, 15 years later.

At that time, Kurt Nielsen was fully engaged in his PhD when he got into a conversation with Professor Ivan Damgård. Over the following years, they worked closely on developing and deploying the new encryption technology, and a collaboration with the food producer Danisco became their major breakthrough.

Danisco, undergoing extensive restructuring at the time, became the first company to adopt the new encryption technology. The collaboration was the industry breakthrough that the research team had been working towards for years – both technologically and commercially. From then, things gained momentum, and more public and private collaborations followed.

From technological breakthrough to entrepreneurial adventure

From thinking you have the solution to creating a viable business is a long journey, says Kurt Nielsen. Partisia was an early adopter of a completely new technology, and it took many years before the market was ready to embrace it, he explains.

Along the way, one encounters a lot of resistance and difficult discussions about everything from strategy to finance. One must be prepared for that, and to have a chance at success, it requires the team to challenge each other, listen, and compromise when necessary.

What does it mean to have a background as a researcher when establishing a company?

“In my work as a researcher and lecturer at the University of Copenhagen, new ideas constantly emerge. I am driven by these ideas and gain energy from them. I believe that many researchers have the same driving force—the key to success with a product or a company lies in the commercial approach, and it varies greatly from researcher to researcher.

Not everyone has it, but there are incredibly talented basic researchers who have a good understanding of how to take an idea forward and who understand that it requires collaboration with people with different expertise to advance one’s idea.

Personally, throughout my university career, I have always been an ‘entrepreneur,’ and early in my career, I helped some friends start a company. However, the dot-com bubble burst, and I returned to the university for a PhD. while considering my future options.

Being an entrepreneur is a state of mind. You must constantly look for new opportunities and be interested in assembling strong teams.

This applies internally in a company and at the university when working together on research projects. The team is crucial.”

Where does this entrepreneurial spirit come from?

“That’s a good question. I have always sought out opportunities, trying to create something through my work and not be locked into a specific job for a lifetime. In reality, I have never considered myself an employee, even though I have received and continue to receive a salary as a researcher. It is the desire to create and develop that drives me.”

About Partisia
Partisia is a spinout from Aarhus University established by internationally renowned researchers and experts in advanced cryptography, business economics, and software development, with experience bringing research ideas to market. The combination of skills enables Partisia to deliver solutions that are both robust and highly innovative in a timely manner.

As a pioneer, Partisia has been selling secure multiparty computation (MPC) and other software solutions for privacy protection since 2008. Initially focusing on secure auctions for commodities such as production contracts, energy-related products, and auctions used for the sale of spectrum licenses.

Since the first commercial use, MPC technology has matured significantly, becoming more agile and notably faster, gradually transforming MPC into a generic infrastructure for privacy-preserving computations.

In parallel with this development, Partisia has developed infrastructure for managing encryption keys and a generic infrastructure for secure computation, as well as various applications across platforms from cloud computing to blockchain technologies.

As part of the commercialization strategy, selected companies have been moved and matured into separate spinouts alongside investors and other business partners, namely Sepior.com (cryptographic key management) and Partisia (applications and infrastructure combining MPC and blockchain technologies). In 2022, Sepior was sold to a major American blockchain company, Blockdaemon, and Partisia is now actively scaling up, focusing on quantum computers.

Read about DIREC’s focus area Digital Tech Startups

The Partisia team assembled in the summer of 2023

Categories
Previous events

PhD defence: Hao Miao

PhD Defence by Hao Miao

Aspects of Deep Spatio-Temporal Analytics for Time Series, Streaming, and Trajectory Data

Abstract

The widespread deployment of wireless and mobile devices results in a proliferation of spatio- temporal data (e.g., time series, streaming, and trajectory data) that is essential for applications, e.g., traffic and air quality prediction, where spatio-temporal analytics plays a key role in ensuring safety, predictability, and reliability. While recent studies have demonstrated superior performance in deep spatio-temporal analytics, many approaches struggle to adapt to real-world conditions. In particular, existing methods suffer from three limitations: 1) existing deep methods typically require large-scale training data, incurring high storage and computational costs; 2) when applied to streaming data, many models suffer from catastrophic forgetting, resulting in deteriorating prediction accuracy over time; and 3) existing solutions often assume centralized data, which leads to privacy concerns and fails to exploit decentralized data processing.

This Ph.D. study aims to systematically study deep spatio-temporal analytics with emerging techniques. Specifically, we target four types of functionality: time series dataset condensation, streaming spatio-temporal prediction, federated trajectory recovery, and federated trajectory similarity learning.

First, we address the problem of time series dataset condensation. The goal is to reduce training costs by summarizing large datasets into smaller, synthetic datasets that can then be used for training instead. We introduce TimeDC, an efficient time series dataset condensation framework that uses two-fold modal matching. TimeDC encompasses a time series feature extraction module for effective feature learning, a decomposition-driven frequency matching module for achieving temporal correlation preservation, and a curriculum- based trajectory matching module for ensuring that the synthetic datasets capture key patterns in the original dataset.

Second, we investigate spatio-temporal prediction on streaming data. We propose URCL, a unified replay-based streaming framework with three key modules: data integration, spatio-temporal continuous representation learning, and spatio-temporal prediction. Specifically, a spatio-temporal mixup mechanism is introduced to alleviate catastrophic forgetting, and a simple siamese network is designed to facilitate holistic feature learning.

Third, we study the problem of federated trajectory recovery, focusing on privacy preservation and enabling decentralized training. We propose LighTR+, a horizontal federated framework, which consists of a lightweight local trajectory embedding module, an intra-domain knowledge distillation module, and a meta-knowledge enhanced local-global training scheme. LighTR+ alleviates intra- and inter-domain interferences across clients while reducing communication costs between clients and the server, thereby facilitating privacy protection and improving efficiency.

Fourth, we explore federated trajectory similarity learning for decentralized data processing. We propose the F-TSL framework based on horizontal federated learning, a server-clients architecture. The framework includes a local trajectory preprocessing and learning module for clients, a privacy-preserving clustering module, and a server-side aggregation module, where the privacy-preserving clustering module leverages differential privacy to handle data heterogeneity across clients.

See event here

 

 
 
 
Categories
News

ChatGPT ushers research into a new era – calling for new rules of the game

1 December 2023

ChatGPT ushers research into a new era – calling for new rules of the game

Language models like ChatGPT will change how we conduct research. To ensure transparency and maintain trust in research, a code of conduct should be drawn up, writes the Royal Academy in Altinget, an online Danish political review.

Authors of the opinion piece:

  • Kim Guldstrand Larsen, professor, Department of Computer Science, Aalborg University, member of Royal Academy
  • Susanne Ditlevsen, chair of the natural sciences class at the Royal Academy and professor at the Department of Mathematical Sciences, University of Copenhagen
  • Thomas Sinkjær, secretary general, Royal Academy and professor at the Department of Medicine and Health Technology, Aalborg University
  • Kristoffer Frøkjær, head of communications, Royal Academy

Israeli researchers recently produced a research paper in less than an hour with the help of ChatGPT.

Fluent and insightful, the article adhered to the expected format of scientific articles. The results, however, were nowhere near close to being novel. This will change in the future.

Read the Altinget opinion piece