• DARIAH EU
• 2013-2022close
• Publications close
• Research data close
• Article close
• DARIAH EU close

A mainstream tendency for multimediality of information and exponential growth of data have promoted major changes in the academic paradigm, supporting the transition towards interdisciplinary research-based studies, which has affected the development of the emerging field of Digital Humanities (DH). The paper aims at providing insights into the different activities implemented by the DH community in Latvia, which have contributed to the advancement of the field and stimulated the inclusion of digital research-related courses in the Latvian higher education curricula, reflecting on the establishment and implementation of the Master study programme in Digital Humanities at RTU considering it in its wider Latvian, pan-European, and global context.

Gazetteers, i.e., lists of place-names, enable having a global vision of places of interest through the assignment of a point, or a region, to a place name. However, such identification of the location corresponding to a place name is often a difficult task. There is no one-to-one correspondence between the two sets, places and names, because of name variants, different names for the same place and homonymy; the location corresponding to a place name may vary in time, changing its extension or even the position; and, in general, there is the imprecision deriving from the association of a concept belonging to language (the place name) to a precise concept (the spatial location). Also for named time periods, e.g., early Bronze Age, which are of current use in archaeology, the situation is similar: they depend on the location to which they refer as the same period may have different time-spans in different locations. The present paper avails of a recent extension of the CIDOC CRM called CRMgeo, which embeds events in a spatio-temporal 4-dimensional framework. The paper uses concepts from CRMgeo and introduces extensions to model gazetteers and period thesauri. This approach enables dealing with time-varying location appellations as well as with space-varying period appellations on a robust basis. For this purpose a refinement/extension of CRMgeo is proposed and a discretization of space and time is used to approximate real space---time extents occupied by events. Such an approach solves the problem and suggests further investigations in various directions.

Exploring the connections between successive phases and overlapping layers from different ages in an ancient building is paramount for its understanding and study. Archaeologists and cultural heritage experts are always eager to unveil the hidden relations of an archaeological building to reconstruct its history and for its interpretation. This paper presents CRMba, a CIDOC CRM extension developed to facilitate the discovery and the interpretation of archaeological resources through the definition of new concepts required to describe the complexity of historic buildings. The CRMba contributes to solving the datasets interoperability issue by exploiting the use of the CIDOC CRM to overcome data fragmentation, to investigate the semantics of building components, of functional spaces and of the construction phases of historic buildings and complexes, making explicit their physical and topological relations through time and space. The approach used for the development of the CRMba makes the model valid for the documentation of different kinds of buildings, across periods, styles and conservation state.

Abstract‘Social and Cultural Innovation’ is a syntagma that is receiving increased usage among researchers since it was the title chosen by the European Strategy Forum Research Infrastructures for the working group that deals with research infrastructures primarily connected with Social Sciences and the Humanities. Innovation refers to the creation of new products and services by bringing a new idea to the market. Economic growth turns on infrastructures, which provide access to services and knowledge, e.g. by overcoming the digital divide. The current migrant and refugee crisis has made it clear with extraordinary effectiveness that a most urgent objective is to work out policies of social and cultural innovation to the advantage of new citizens – policies that will make them feel welcome in full dignity.

This article shows that metadata plays a central role in our society and concludes that through collaborative work, it is possible to pool solutions and to establish relationships of cooperation, both at the level of practical tool development and with regard to sharing and creating knowledge and know-how. ispartof: ABB: Archives et Bibliothèques de Belgique - Archief- en Bibliotheekwezen in België vol:106 pages:135-144 status: published

Obtaining high-quality labelled data for training a classifier in a new application domain is often costly. Transfer Learning (a.k.a. “Inductive Transfer”) tries to alleviate these costs by transferring, to the “target” domain of interest, knowledge available from a different “source” domain. In transfer learning the lack of labelled information from the target domain is compensated by the availability at training time of a set of unlabelled examples from the target distribution. Transductive Transfer Learning denotes the transfer learning setting in which the only set of target documents that we are interested in classifying is known and available at training time. Although this definition is indeed in line with Vapnik’s original definition of “transduction”, current terminology in the field is confused. In this article, we discuss how the term “transduction” has been misused in the transfer learning literature, and propose a clarification consistent with the original characterization of this term given by Vapnik. We go on to observe that the above terminology misuse has brought about misleading experimental comparisons, with inductive transfer learning methods that have been incorrectly compared with transductive transfer learning methods. We then, give empirical evidence that the difference in performance between the inductive version and the transductive version of a transfer learning method can indeed be statistically significant (i.e., that knowing at training time the only data one needs to classify indeed gives an advantage). Our clarification allows a reassessment of the field, and of the relative merits of the major, state-of-the-art algorithms for transfer learning in text classification.