DISA

When? March 13, 14.00-16.00
Where? Onsite: D2272 and via zoom
Registration: Please sign up for the seminar via this link https://forms.gle/EbRUASvY9c73kMNz6 by March 11. This is especially important if you plan to attend onsite so we can make sure there is fika for everyone

Agenda
14.00-14.10 Welcome and practical information
14.10-14.55 Palme archives as a chatbot – Tibo Bruneel
14.55 – 15.05 Coffee break
15.05 – 15.50 Gradient Tree Boosting for Regression Transfer – Dag Björnberg
15.50 -16.00 Sum up and plan for our upcoming seminars

Abstracts

Palme archives as a chatbot – Tibo Bruneel  
For nearly four decades, the assassination of Swedish Prime Minister Olof Palme has remained a complex and heavily disputed case. Although the investigation officially closed in 2020, the search for answers continues within a monumental digital footprint. This dataset is a chaotic, unstructured web of typed police reports, handwritten notes, maps, and images. 

How can we potentially find previously unknown clues buried within decades of scattered data? 
This presentation introduces “PalmeNet-Chat,” an LLM-powered investigative tool developed by Softwerk AB in collaboration with the true-crime podcast Spår. We will detail the technical challenge of processing this dense archive. By engineering an on-premise OCR pipeline utilising Vision Language Models, we transformed raw history into a structured, searchable library. We will then explore how we implemented Retrieval Augmented Generation (RAG) and vector databases to build a system capable of semantic search and contextual reasoning across the entire case file. Finally, we will offer a glimpse into the next phase of our project, showcasing how we are taking this investigative tool to an entirely new level. 

Gradient Tree Boosting for Regression Transfer – Dag Björnberg  
Many real-world modeling problems are hindered by limited data availability. In such cases, transfer learning leverages related source domains to improve predictions in a target domain of interest. We extend the classical gradient tree boosting paradigm to a regression transfer algorithm by modeling the weak learner as a sum of two regression trees. The trees are fitted on source data and target data, respectively, and jointly optimized for the target data. We derive optimal coefficients for the model update under the least-squares, the least-absolute deviation, and the Huber loss functions. We benchmark our approach against the widely used XGBoost algorithm in several transfer scenarios, achieving superior performance in seven out of eight cases. 

Where? Onsite: D2272 and via zoom
Registration: Please sign up for the seminar via this link https://forms.gle/QB5oiSWMpVofjBHY6 by February 11. This is especially important if you plan to attend onsite so we can make sure there is fika for everyone.

Abstracts
Novelty Detection Using Time-Series Transient Data from the Lead-Acid Forklift Batteries – Zijie Feng  
In industrial applications, monitoring the battery health of electrically powered machinery and detecting abnormal operating conditions is a persistent yet critical challenge. Traditionally, anomaly detection systems are developed reactively: abnormalities are identified only after they occur, often leading to operational disruptions and economic losses. Novelty detection offers an alternative by learning normal behavior and detecting previously unseen abnormalities.  

In this work, we compare a diverse set of data-driven novelty detection methods using simulated time-series transient data derived from real lead-acid forklift battery measurements, aiming to identify suitable solutions for different types of anomalies. 

Potential of Graph Neural Networks for Software Architecture Recovery– Rakshanda Jabeen  
Software architecture recovery (SAR) aims to uncover a system’s modular structure directly from source code, supporting comprehension and maintenance when documentation is missing or outdated. In this work, we investigate the potential of graph neural networks (GNNs) and unsupervised learning for SAR by modeling software systems as heterogeneous, multi-relational graphs. Nodes represent software entities (e.g., files or classes) and typed edges capture structural and functional dependencies such as calls, imports, inheritance, and other code-level relations. To complement dependency structures with meaning, we integrate semantic signals from source code identifiers and related textual artifacts via contextual code embeddings (e.g., Word2Vec), yielding representations that capture both what entities do and how they interact. 

We study heterogeneous GNN encoders that aggregate information across relation types, including heterogeneous graph convolution and heterogeneous attention mechanisms, to analyze the trade-off between fixed normalization and adaptive neighbor weighting in software graphs. On top of these encoders, we explore two unsupervised training paradigms: (i) graph autoencoding, where embeddings are learned by reconstructing observed dependency relations, and (ii) contrastive representation learning inspired by Deep Graph Infomax, which maximizes agreement between embeddings from the original graph and perturbed views. The resulting entity embeddings are clustered to recover candidate architectural modules. Preliminary results across multiple open-source systems indicate that combining semantic cues with structural and functional dependencies produces more meaningful module separation than using structure alone, demonstrating that modern graph representation learning is a promising direction for robust, automated SAR beyond heuristic baselines.