727 research outputs found
Sort by
What makes up a reportable event in a language? Motion events as an important test domain in linguistic typology [Dataset]
No full textNumerous cross-linguistic studies on motion events have been carried out in investigating the scope of the two-fold typology path versus manner (Talmy 1985; 2000) and its possible implications. This typological contrast is too narrow as it stands, however, to account for the diversity found within and across types. The present study is based on what can be termed a ‘process-oriented perspective’. It includes the analyses of all relevant conceptual domains notably the domain of temporality, in addition to space, and thus goes beyond previous studies. The languages studied differ typologically as follows: ‘path’ is typically expressed in the verb in French and Tunisian Arabic vs ‘manner of motion’ in English and German, while in the temporal domain ‘aspect’ is expressed in English and Tunisian Arabic but not in German and French. The study thus compares the representations which speakers construct when forming a reportable event as a response to video clips of a series of naturalistic scenes in which an entity moves through space. The analysis includes the following conceptual categories: (1) the privileged event ‘layer’ (manner versus path) which drives the selection of breakpoints for forming event units when processing the visual input, (2) the privileged category in spatial framing (figure-based/ground-based), and (3) viewpoint aspect (phasal decomposition or not). We assume that each of these three cognitive categories is shaped specifically by language structure (system and repertoire) and language use (frequency of constructions). The findings reveal systematic differences both across, as well as within, typologically-related languages with respect to (1) the basic event type encoded, (2) the changes in quality expressed, (3) the total number of path segments encoded per situation, and (4) the number of path segments packaged into one utterance. The findings reveal what can be termed language-specific default settings along each of the conceptual dimensions and their interrelations which function as ‘language specific attentional templates’
Late Antique Statue Bases of Lepcis Magna
No full textInteractive and searchable map of the city of Lepcis Magna, displaying the geographical distribution within the city scape of the late antique honorary and building inscriptions.
When translating a multi-faceted and multi-layered reality such as that of late antique Lepcis into a categorized environment, a certain level of simplification is needed, therefore reference to the written sources for detailed discussion remains indispensable (see I. Tantillo / F. Bigi, Leptis Magna. Una città e le sue iscrizioni in epoca tardoromana, Cassino 2010; as the map itself is a simplified rendering of the city and its buildings, please refer to existing bibliography for detailed maps, e.g. R. Bianchi Bandinelli / E. Vergara Caffarelli / G. Caputo, Leptis Magna, Milano 1964).
The map features all public inscriptions – published so far – produced in Lepcis in the period between 250 and 420 ad (six minor fragments with insufficient data have been left out: HD059617; HD059642; HD059661; HD067591; HD067593; HD059436). It also includes a series of four monuments whose inscription is now lost but which can be nonetheless assigned to the same period (B.L.T. = Bases with Lost Text).
All the relevant data have been collected in a searchable database and grouped in ten fields
Datenpublikation zu materialen Formierungen musiktheoretischer Konzepte: Praxeologie eines Fachschrifttums im ausgehenden Mittelalter
No full textDas Projekt untersucht in der Überlieferung musiktheoretischer Texte an der Wende vom Spätmittelalter zur frühen Neuzeit die inhaltliche Konstitution, materiale Gestaltung und praktische Verwendung von Handschriften (und auch frühen Drucken): Wie wird theoretisches Wissen über Musik materiell präsentiert, um effizient vermittelbar zu sein, und welche Auswirkungen haben die Praktiken der Weitergabe auf die Textträger und Inhalte selbst?
Derartiges Fachschrifttum diente häufig didaktischen Zwecken (universitärem Lehrbetrieb, Klerikerausbildung oder Schulunterricht), so dass sich daran die Mechanismen und Inhalte der disziplinären Wissensvermittlung analysieren lassen, insbesondere der Einsatz von unterschiedlichen medialen Elementen (Schrift, Notentext, Diagramme) und Rezeptionspraktiken im Bereich der Aneignung und Personalisierung des Geschriebenen durch die Akteure (z.B. Lektüre, Annotation, Bearbeitung, Kompilation).
Zur Zeit sind als Pilotprojekt 4 bearbeitete Handschriften bzw. -auszüge zugänglich. Diese Plattform wurde am 8. Dezember 2018 in einer ersten Version freigeschaltet und befindet sich weiterhin im Aufbau.
Die Bearbeitung der zuerst bereitgestellten Handschriftenbeispiele konzentriert sich auf Diagramme. Sie sind ein essentieller Teil der Musiklehre. Wie in den anderen quadrivialen Disziplinen dienten Diagramme in der Musik zur Vermittlung der Grundlagen, z. B. der Proportionsverhältnisse oder des Tonsystems. Durch den Übertrag von Tonrelationen auf Raumrelationen wurde es möglich etwas sichtbar zu machen, was noch bei Isidor als unmöglich beschrieben wurde: Das Hörbare auf Papier zu bringen. Allzu häufig werden Diagramme trotzdem als Beiwerk zum Text betrachtet und nicht als essentieller Teil der Handschrift gesehen, dabei sind sie in weitverbreiteten Schriften wie der musica enchiriades oder Johannes de Muris‘ Musica speculativa eines der wesentlichen Elemente zum Verständnis der Lehre.
Diagramme haben im Gegensatz zu Texten jedoch einen erheblichen Nachteil: Sie sind nicht sequenziell. Verfügbare Textdatenbanken (s. Thesaurus musicarum latinarum und Lexicon musicum Latinum medii aevi) bieten Möglichkeiten, Texte und Konkordanzen zwischen einzelnen Traktaten zu vergleichen. Der Vergleich von Diagrammen ist allerdings wegen ihrer noch weitgehend ausstehenden systematischen Verzeichnung und des Fehlens etablierter Beschreibungskriterien bislang nur sehr begrenzt möglich.
Diese Plattform bietet die Möglichkeit, Diagramme und Notenbeispiele der im Korpus befindlichen Handschriften zu vergleichen. Diagramme einer oder mehrerer Handschriften lassen sich anhand von Schlagworten sortieren und somit gegenüberstellen. Hieraus lassen sich Verwandtschaften und Entwicklungslinien erkennen. Diagramme werde auf diese Weise aus ihrer isolierten Betrachtung geholt.
Die Verwendung von Diagrammen ist im Wesentlichen abhängig von den materialen Eigenschaften einer Handschrift. Beim Schreiben müssen beispielsweise gezielt Raum gelassen werden, um Diagramme nachtragen zu können. Diagramme und Handschriften erscheinen deshalb als Digitalisate der Originalhandschriften. Erst mit einem Blick auf die ganze Seite lässt sich erschließen, warum eine bestimmte Darstellung gewählt wurde.</p
AMR parse quality prediction [Source Code]
No full textAccuracy prediction for AMR parsing predicts 33 accuracy metrics for a given sentence and its (automatic) AMR parse
Abstract (Opitz and Frank, 2019):
Semantic proto-role labeling (SPRL) is an alternative to semantic role labeling (SRL) that moves beyond a categorical definition of roles, following Dowty's feature-based view of proto-roles. This theory determines agenthood vs. patienthood based on a participant's instantiation of more or less typical agent vs. patient properties, such as, for example, volition in an event. To perform SPRL, we develop an ensemble of hierarchical models with self-attention and concurrently learned predicate-argument-markers. Our method is competitive with the state-of-the art, overall outperforming previous work in two formulations of the task (multi-label and multi-variate Likert scale prediction). In contrast to previous work, our results do not depend on gold argument heads derived from supplementary gold tree banks.
</p
DCE-MRI of the human kidney using BLADE: A feasibility study in healthy volunteers [Dataset]
No full textTo evaluate the degree of motion compensation in the kidney using two different sampling methods, each in their optimized settings: A BLADE k-space acquisition technique and a routinely used kidney perfusion acquisition scheme (TurboFLASH).
Dynamic contrast enhanced magnetic resonance examinations were performed in 16 healthy volunteers on a 3 Tesla MR-system with two parameterizations of the BLADE sequence and the standard reference acquisition scheme. Signal intensity enhanced time curves were analyzed with a mathematical model and a widely published separable compartment model on cortex regions to assess robustness versus motion artifact
Abstract graphs, abstract paths, grounded paths for Freebase and NELL
No full textWe describe a method for representing knowledge graphs that capture an intensional representation of the original extensional information. This representation is very compact, and it abstracts away from individual links, allowing us to find better path candidates, as shown by the results of link prediction using this information.
The data in this archive consists of the abstract graphs built from several configurations of Freebase and NELL (used in the experiments described in Gardner et al. 2014), the abstract paths extracted from these graphs, the grounded paths and the negative sampling used in the experiments described in (Nastase and Kotnis, 2019). The motivation for this work was to find better patterns in knowledge graphs than those obtained using the PRA approach.
This dataset contains:
AbstractGraphs_Garder2014data.tar.gz: the abstract graphs
AbstractPathsGroundedPaths.tar.gz: the abstract paths, grounded paths, train/test data and the corresponding negative samples (for each abstract graph)
For a brief description of the data in each archive, please see README_AbstractGraphs.txt. </p
Multilingual Modal Sense Classification using a Convolutional Neural Network [Source Code]
No full textAbstract
Modal sense classification (MSC) is aspecial WSD task that depends on themeaning of the proposition in the modal’s scope. We explore a CNN architecture for classifying modal sense in English and German. We show that CNNs are superior to manually designed feature-based classifiers and a standard NN classifier. We analyze the feature maps learned by the CNN and identify known and previously unattested linguistic features. We bench-mark the CNN on a standard WSD task,where it compares favorably to models using sense-disambiguated target vectors.
(Marasović and Frank, 2016)</p
Pooled clone collections by multiplexed CRISPR-Cas12a-assisted gene tagging in yeast [Dataset]
No full textData accompanying the paper "Pooled clone collections by multiplexed CRISPR-Cas12a-assisted gene tagging in yeast" by Buchmuller and Herbst et al, 2019, Nat Communications. This contains raw NGS data for all genotyping analysis in the publication as well as the source code of the castR design software (v1.0)
Affixoid Dataset (DE)
No full textThe dataset contains the manual annotations for the COLING 2018
submission "Distinguishing affixoid formations from compounds" by Josef
Ruppenhofer, Michael Wiegand, Rebecca Wilm and Katja Markert.
1788 complex words containing one of 7 German suffixoid candidates (e.g.
-hai, -gott) were annotated manually as to whether the complex forms
represent regular compounds or affixoid formations. The main experiments
in the paper use automatically extracted features of the complex forms
in trying to correctly make this distinction.
Additionally, the words were labeled for five properties related to any
intensifying and evaluative meaning potentially associated with the
whole word and its components. These manual feature annotations were
used to establish the upper-bound performance of a classifier trained
to distinguish affixoid formations from regular compounds
Fully-automated quality assurance in multi-center studies using MRI phantom measurements [Dataset]
No full text43 measurements acquired in eight different sites within the
IMAGEN-project, comprising the following 3 T scanner types: Siemens Verio and TimTrio; General Electric Signa Excite, and Signa HDx; Philips Achieva.
Additionally one phantom data set was aquired on a 3 T Siemens Skyra resulting in 44 measurements