Author name cluster

Karl Aberer

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

8 papers

2 author rows

ECAI Conference 2025 Conference Paper

LoRA-XS: Low-Rank Adaptation with Extremely Small Number of Parameters

Klaudia Balazy
Mohammadreza Banaei
Karl Aberer
Jacek Tabor

The growth of large language models underscores the need for parameter-efficient fine-tuning. Despite its popularity, LoRA encounters storage and computational challenges when deploying multiple task- or user-specific modules. To address this, we introduce LoRA-XS, a novel fine-tuning method backed by a theoretical derivation. LoRA-XS drastically reduces trainable parameters by incorporating a small, trainable weight matrix between frozen low-rank matrices derived from the Singular Value Decomposition of pre-trained weights. This design enables LoRA-XS to reduce storage requirements by over 100x in 7B models compared to LoRA. Additionally, unlike other methods, LoRA-XS imposes no lower bound on trainable parameters – it can scale from a single parameter per module to arbitrarily large values, adapting to any storage or computational constraint. Evaluations on GLUE, GSM8K, MATH, and commonsense reasoning benchmarks across different model scales reveal that LoRA-XS consistently outperforms or matches LoRA and VeRA in accuracy, offering unmatched parameter efficiency. Our ablation studies highlight the significance of singular vectors in transformer weights, establishing LoRA-XS as a powerful, storage-efficient solution for scaling and personalizing large language models.

Details

AAAI Conference 2018 Conference Paper

280 Birds With One Stone: Inducing Multilingual Taxonomies From Wikipedia Using Character-Level Classification

Amit Gupta
Rémi Lebret
Hamza Harkous
Karl Aberer

We propose a novel fully-automated approach towards inducing multilingual taxonomies from Wikipedia. Given an English taxonomy, our approach ﬁrst leverages the interlanguage links of Wikipedia to automatically construct training datasets for the is-a relation in the target language. Character-level classiﬁers are trained on the constructed datasets, and used in an optimal path discovery framework to induce high-precision, high-coverage taxonomies in other languages. Through experiments, we demonstrate that our approach signiﬁcantly outperforms the state-of-the-art, heuristics-heavy approaches for six languages. As a consequence of our work, we release presumably the largest and the most accurate multilingual taxonomic resource spanning over 280 languages.

PDF Details

IJCAI Conference 2017 Conference Paper

Hybrid Neural Networks for Learning the Trend in Time Series

Tao Lin
Tian Guo
Karl Aberer

The trend of time series characterizes the intermediate upward and downward behaviour of time series. Learning and forecasting the trend in time series data play an important role in many real applications, ranging from resource allocation in data centers, load schedule in smart grid, and so on. Inspired by the recent successes of neural networks, in this paper we propose TreNet, a novel end-to-end hybrid neural network to learn local and global contextual features for predicting the trend of time series. TreNet leverages convolutional neural networks (CNNs) to extract salient features from local raw data of time series. Meanwhile, considering the long-range dependency existing in the sequence of historical trends of time series, TreNet uses a long-short term memory recurrent neural network (LSTM) to capture such dependency. Then, a feature fusion layer is to learn joint representation for predicting the trend. TreNet demonstrates its effectiveness by outperforming CNN, LSTM, the cascade of CNN and LSTM, Hidden Markov Model based method and various kernel based baselines on real datasets.

PDF Details

AAMAS Conference 2013 Conference Paper

An MAS Negotiation Support Tool for Schema Matching

Hung Quoc Viet Nguyen
Xuan Hoai Luong
Zoltán Miklós
Tho Quan Thanh
Karl Aberer

Database schema matching is the process of establishing correspondences between attributes of schemas for data integration purpose. Though various commercial tools have been developed, their results are inherently uncertain. In practice, to obtain correct attribute correspondences, there is a need for collecting human input, after the use of automatic matching tools, to reconcile erroneous mappings. We present a negotiation support tool that enables not a single expert but an expert team, whose members might have conflicting views, can work collaboratively to reconcile the output of the automatic tools. In an attempt to facilitate and support cooperation in team integration, our tool sets the goal to compute all possible decisions from expert inputs as well as explanations for each decision. Moreover, it also shows the foreseeable consequences of choosing a particular decision. Technically, this tool is developed on top of an argumentation framework.

PDF

TIST Journal 2013 Journal Article

Semantic trajectories

Zhixian Yan
Dipanjan Chakraborty
Christine Parent
Stefano Spaccapietra
Karl Aberer

With the large-scale adoption of GPS equipped mobile sensing devices, positional data generated by moving objects (e.g., vehicles, people, animals) are being easily collected. Such data are typically modeled as streams of spatio-temporal (x,y,t) points, called trajectories. In recent years trajectory management research has progressed significantly towards efficient storage and indexing techniques, as well as suitable knowledge discovery. These works focused on the geometric aspect of the raw mobility data. We are now witnessing a growing demand in several application sectors (e.g., from shipment tracking to geo-social networks) on understanding the semantic behavior of moving objects. Semantic behavior refers to the use of semantic abstractions of the raw mobility data, including not only geometric patterns but also knowledge extracted jointly from the mobility data and the underlying geographic and application domains information. The core contribution of this article lies in a semantic model and a computation and annotation platform for developing a semantic approach that progressively transforms the raw mobility data into semantic trajectories enriched with segmentations and annotations. We also analyze a number of experiments we did with semantic trajectories in different domains.

Details DOI

TAAS Journal 2011 Journal Article

Effective Usage of Computational Trust Models in Rational Environments

Le-Hung Vu
Karl Aberer

Computational reputation-based trust models using statistical learning have been intensively studied for distributed systems where peers behave maliciously. However practical applications of such models in environments with both malicious and rational behaviors are still very little understood. In this article, we study the relation between their accuracy measures and their ability to enforce cooperation among participants and discourage selfish behaviors. We provide theoretical results that show the conditions under which cooperation emerges when using computational trust models with a given accuracy, and how cooperation can still be sustained while reducing the cost and accuracy of those models. Specifically, we propose a peer selection protocol that uses a computational trust model as a dishonesty detector to filter out unfair ratings. We prove that such a model with reasonable misclassification error bound in identifying malicious ratings can effectively build trust and cooperation in the system, considering rationality of participants. These results reveal two interesting observations. First, the key to the success of a reputation system in a rational environment is not a sophisticated trust-learning mechanism, but an effective identity-management scheme to prevent whitewashing behaviors. Second, given an appropriate identity-management mechanism, a reputation-based trust model with a moderate accuracy bound can be used to effectively enforce cooperation in systems with both rational and malicious participants. As a result, in heterogeneous environments where peers use different algorithms to detect misbehavior of potential partners, cooperation may still emerge. We verify and extend these theoretical results to a variety of settings involving honest, malicious, and strategic players through extensive simulation. These results will enable a much more targeted, cost-effective and realistic design for decentralized trust management systems, such as needed for peer-to-peer, electronic commerce, or community systems.

Details DOI

TAAS Journal 2010 Journal Article

Structured overlay for heterogeneous environments

Šarūnas Girdzijauskas
Anwitaman Datta
Karl Aberer

Recent years have seen advances in building large Internet-scale index structures, generally known as structured overlays. Early structured overlays realized distributed hash tables (DHTs) which are ill suited for anything but exact queries. The need to support range queries necessitates systems that can handle uneven load distributions. However such systems suffer from practical problems—including poor latency, disproportionate bandwidth usage at participating peers, or unrealistic assumptions on peers' homogeneity, in terms of available storage or bandwidth resources. In this article we consider a system that is not only able to support uneven load distributions but also to operate in heterogeneous environments, where each peer can autonomously decide how much of its resources to contribute to the system. We provide the theoretical foundations of realizing such a network and present a newly proposed system Oscar based on these principles. Oscar can construct efficient overlays given arbitrary load distributions by employing a novel scalable network sampling technique. The simulations of our system validate the theory and evaluate Oscar's performance under typical challenges, encountered in real-life large-scale networked systems, including participant heterogeneity, faults, and skewed and dynamic load-distributions. Thus the Oscar distributed index fills in an important gap in the family of structured overlays, bringing into life a practical Internet-scale index, which can play a crucial role in enabling data-oriented applications distributed over wide-area networks.

Details DOI

EAAI Journal 2005 Journal Article

Probabilistic prediction of peers’ performance in P2P networks

Zoran Despotovic
Karl Aberer

The problem of encouraging trustworthy behavior in P2P online communities by managing peers’ reputations has drawn a lot of attention recently. However, most of the proposed solutions exhibit the following two problems: huge implementation overhead and unclear trust related model semantics. This paper shows that a simple probabilistic technique, maximum likelihood estimation namely, can reduce these two problems substantially when employed as the feedback aggregation strategy. We evaluate the technique in three settings relevant for applications of P2P networks and show that it performs well in all of them. Thus, no complex exploration of the feedback is necessary. Instead, simple, intuitive and efficient probabilistic estimation methods suffice.

Details DOI