arXiv subjects

Michael Fromm

Publications and source records attributed to Michael Fromm.

30 records · Page 2Linked to original sources

arXiv2021

Diversity Aware Relevance Learning for Argument Search

In this work, we focus on the problem of retrieving relevant arguments for a query claim covering diverse aspects. State-of-the-art methods rely on explicit mappings between claims and premises, and thus are unable to utilize large available collections of premises without laborious and costly manual annotation. Their diversity approach relies on removing duplicates via clustering which does not directly ensure that the selected premises cover all aspects. This work introduces a new multi-step approach for the argument retrieval problem. Rather than relying on ground-truth assignments, our approach employs a machine learning model to capture semantic relationships between arguments. Beyond that, it aims to cover diverse facets of the query, instead of trying to identify duplicates explicitly. Our empirical evaluation demonstrates that our approach leads to a significant improvement in the argument retrieval task even though it requires less data.

cs.IR↗

arXiv2020

Argument Mining Driven Analysis of Peer-Reviews

Peer reviewing is a central process in modern research and essential for ensuring high quality and reliability of published work. At the same time, it is a time-consuming process and increasing interest in emerging fields often results in a high review workload, especially for senior researchers in this area. How to cope with this problem is an open question and it is vividly discussed across all major conferences. In this work, we propose an Argument Mining based approach for the assistance of editors, meta-reviewers, and reviewers. We demonstrate that the decision process in the field of scientific publications is driven by arguments and automatic argument identification is helpful in various use-cases. One of our findings is that arguments used in the peer-review process differ from arguments in other domains making the transfer of pre-trained models difficult. Therefore, we provide the community with a new peer-review dataset from different computer science conferences with annotated arguments. In our extensive empirical evaluation, we show that Argument Mining can be used to efficiently extract the most relevant parts from reviews, which are paramount for the publication decision. The process remains interpretable since the extracted arguments can be highlighted in a review without detaching them from their context.

cs.CY↗

arXiv2020

Unsupervised Anomaly Detection for X-Ray Images

Obtaining labels for medical (image) data requires scarce and expensive experts. Moreover, due to ambiguous symptoms, single images rarely suffice to correctly diagnose a medical condition. Instead, it often requires to take additional background information such as the patient's medical history or test results into account. Hence, instead of focusing on uninterpretable black-box systems delivering an uncertain final diagnosis in an end-to-end-fashion, we investigate how unsupervised methods trained on images without anomalies can be used to assist doctors in evaluating X-ray images of hands. Our method increases the efficiency of making a diagnosis and reduces the risk of missing important regions. Therefore, we adopt state-of-the-art approaches for unsupervised learning to detect anomalies and show how the outputs of these methods can be explained. To reduce the effect of noise, which often can be mistaken for an anomaly, we introduce a powerful preprocessing pipeline. We provide an extensive evaluation of different approaches and demonstrate empirically that even without labels it is possible to achieve satisfying results on a real-world dataset of X-ray images of hands. We also evaluate the importance of preprocessing and one of our main findings is that without it, most of our approaches perform not better than random. To foster reproducibility and accelerate research we make our code publicly available at https://github.com/Valentyn1997/xray

eess.IV↗

arXiv2019

Using Supervised Learning to Classify Metadata of Research Data by Discipline of Research

Automated classification of metadata of research data by their discipline(s) of research can be used in scientometric research, by repository service providers, and in the context of research data aggregation services. Openly available metadata of the DataCite index for research data were used to compile a large training and evaluation set comprised of 609,524 records, which is published alongside this paper. These data allow to reproducibly assess classification approaches, such as tree-based models and neural networks. According to our experiments with 20 base classes (multi-label classification), multi-layer perceptron models perform best with a f1-macro score of 0.760 closely followed by Long Short-Term Memory models (f1-macro score of 0.755). A possible application of the trained classification models is the quantitative analysis of trends towards interdisciplinarity of digital scholarly output or the characterization of growth patterns of research data, stratified by discipline of research. Both applications perform at scale with the proposed models which are available for re-use.

cs.IR↗

arXiv2019

TACAM: Topic And Context Aware Argument Mining

In this work we address the problem of argument search. The purpose of argument search is the distillation of pro and contra arguments for requested topics from large text corpora. In previous works, the usual approach is to use a standard search engine to extract text parts which are relevant to the given topic and subsequently use an argument recognition algorithm to select arguments from them. The main challenge in the argument recognition task, which is also known as argument mining, is that often sentences containing arguments are structurally similar to purely informative sentences without any stance about the topic. In fact, they only differ semantically. Most approaches use topic or search term information only for the first search step and therefore assume that arguments can be classified independently of a topic. We argue that topic information is crucial for argument mining, since the topic defines the semantic context of an argument. Precisely, we propose different models for the classification of arguments, which take information about a topic of an argument into account. Moreover, to enrich the context of a topic and to let models understand the context of the potential argument better, we integrate information from different external sources such as Knowledge Graphs or pre-trained NLP models. Our evaluation shows that considering topic information, especially in connection with external information, provides a significant performance boost for the argument mining task.

cs.CL↗

arXiv2013

Onset Transition to Cold Nuclear Matter from Lattice QCD with Heavy Quarks

Lattice QCD at finite density suffers from a severe sign problem, which has so far prohibited simulations of the cold and dense regime. Here we study the onset of nuclear matter employing a three-dimensional effective theory derived by combined strong coupling and hopping expansions, which is valid for heavy but dynamical quarks and has a mild sign problem only. Its numerical evaluations agree between a standard Metropolis and complex Langevin algorithm, where the latter is free of the sign problem. Our continuum extrapolated data clearly show a first order phase transition building up at $μ_B \approx m_B$ as the temperature approaches zero. An excellent description of the data is achieved by an analytic solution in the strong coupling limit.

hep-lat↗

arXiv2012

Phase transitions in heavy-quark QCD from an effective theory

With combined hopping parameter and strong coupling expansions, we calculate a dimensionally reduced Polyakov-loop effective theory valid for heavy quarks at nonzero temperature and arbitrary chemical potential. We numerically compute the critical endpoint of the deconfinement transition as a function of quark masses and number of flavours. We also investigate the applicability of the model to the low-T and high density region, specifically in terms of baryon condensation phenomena.

hep-lat↗

arXiv2011

Strong coupling effective theory with heavy fermions

We extend the recently developed strong coupling, dimensionally reduced Polyakov-loop effective theory from finite-temperature pure Yang-Mills to include heavy fermions and nonzero chemical potential by means of a hopping parameter expansion. Numerical simulation is employed to investigate the weakening of the deconfinement transition as a function of the quark mass. The tractability of the sign problem in this model is exploited to locate the critical surface in the (M/T, mu/T, T) space over the whole range of chemical potentials from zero up to infinity.

hep-lat↗

arXiv2011

The QCD deconfinement transition for heavy quarks and all baryon chemical potentials

Using combined strong coupling and hopping parameter expansions, we derive an effective three-dimensional theory from thermal lattice QCD with heavy Wilson quarks. The theory depends on traced Polyakov loops only and correctly reflects the centre symmetry of the pure gauge sector as well as its breaking by finite mass quarks. It is valid up to certain orders in the lattice gauge coupling and hopping parameter, which can be systematically improved. To its current order it is controlled for lattices up to N_τ\sim 6 at finite temperature. For nonzero quark chemical potentials, the effective theory has a fermionic sign problem which is mild enough to carry out simulations up to large chemical potentials. Moreover, by going to a flux representation of the partition function, the sign problem can be solved. As an application, we determine the deconfinement transition and its critical end point as a function of quark mass and all chemical potentials.

hep-lat↗

arXiv2011

Towards corrections to the strong coupling limit of staggered lattice QCD

We report on the first steps of an ongoing project to add gauge observables and gauge corrections to the well-studied strong coupling limit of staggered lattice QCD, which has been shown earlier to be amenable to numerical simulations by the worm algorithm in the chiral limit and at finite density. Here we show how to evaluate the expectation value of the Polyakov loop in the framework of the strong coupling limit at finite temperature, allowing to study confinement properties along with those of chiral symmetry breaking. We find the Polyakov loop to rise smoothly, thus signalling deconfinement. The non-analytic nature of the chiral phase transition is reflected in the derivative of the Polyakov loop. We also discuss how to construct an effective theory for non-zero lattice coupling, which is valid to $O(β)$.

hep-lat↗

arXiv2010

Nuclear Physics from lattice QCD at strong coupling

We study numerically the strong coupling limit of lattice QCD with one flavor of massless staggered quarks. We determine the complete phase diagram as a function of temperature and chemical potential, including a tricritical point. We clarify the nature of the low temperature dense phase, which is strongly bound nuclear matter. This strong binding is explained by the nuclear potential, which we measure. Finally, we determine, from this first-principle limiting case of QCD, the masses of atomic nuclei up to A=12 "carbon".

hep-lat↗

arXiv2009

Nuclear physics from strong coupling QCD

The strong coupling limit (beta_gauge = 0) of QCD offers a number of remarkable research possibilities, of course at the price of large lattice artifacts. Here, we determine the complete phase diagram as a function of temperature T and baryon chemical potential mu_B, for one flavor of staggered fermions in the chiral limit, with emphasis on the determination of a tricritical point and on the T ~ 0 transition to nuclear matter. The latter is known to happen for mu_B substantially below the baryon mass, indicating strong nuclear interactions in QCD at infinite gauge coupling. This leads us to studying the properties of nuclear matter from first principles. We determine the nucleon-nucleon potential in the strong coupling limit, as well as masses m_A of nuclei as a function of their atomic number A. Finally, we clarify the origin of nuclear interactions at strong coupling, which turns out to be a steric effect.

hep-lat↗