Abstract. In this paper we discuss a cross-cultural analysis of non-verbal expressions (gestures, facial expressions, body posture) that have a turn managing function in the flow of  interaction. The study was carried out by analyzing and comparing the features of interest in two samples of institutional interaction, namely face-to-face political interviews, in Greek and Icelandic respectively. The non-verbal behavior of the participants in both interviews was annotated following the same annotation process. The attested turn management instances were compared in order to find similarities and differences in terms of frequency and modality preference. 

Abstract: The aim of the present work is to investigate the relative contribution of ordered and stochastic components in natural written texts and examine the influence of text category and language on these. To this end, a binary representation of written texts and the generated symbolic sequences are examined by standard block entropy analysis and the Shannon and Kolmogorov entropies are obtained. It is found that both entropies are sensitive to both language and text category with the text category sensitivity following almost the same trends in both languages (English and Greek). The values of these entropies are compared with those of stochastically generated symbolic sequences and the nature of correlations present in this representation of real written texts is identified.

Abstract: In this paper we describe a system that applies emerging technologies for speech recognition, language processing, multimedia indexing and retrieval, all integrated into a large video and audio library that covers broadcast news and current affairs in Greece. It assists the Greek National Council for Radio and Television (NCRTV) in compiling information, annotating and analyzing news and monitoring national, political, social, economic, cultural and environmental issues concerning Greece in general. It further assists supervision of the broadcast A/V sector by offering citizens an efficient way to seek and get hold of, an ‘official’ copy of aired programming, in order to safeguard their interests and promote NCRTV’s goals.

Abstract: This paper describes a web-enabled sophisticated Cultural Heritage (CH) system giving access to digital resources of various media, which exploits Language Technologies (LT) in order to enhance the performance of the search and retrieval mechanisms. More specifically, the paper presents the system requirements and architecture, drawing aspects from: (a) the cultural data repository and its particularities; (b) the unified metadata scheme that has been devised, integrating elements from various metadata standards, providing thus a rich description of the resources; (c) the thesauri (one of the major pillars of the system) that provide uniform access to the resources. The LT that form part of the system construction and use are presented in detail, focusing on the Term Extraction and Named Entity Recognition tools used in the construction of the thesauri and the metadata annotation process, and the Term Matching module exploited in the mining process for the identification of query terms which appear in a morphosyntactically similar form in the thesauri.

Abstract: Throughout the chapter, we provide details on implementation issues of practical systems for efficient multimedia retrieval. Moreover, we exemplify algorithms and technologies by referring to practices and results of an EC-funded project called Combined IMage and WOrd Spotting (CIMWOS) developed with the hope that it would be a powerful tool in facilitating common procedures for intelligent indexing and retrieval of audiovisual material. CIMWOS used a multifaceted approach for the location of important segments within multimedia material employing state-of-the-art algorithms for text, speech and image processing in promoting reuse of audiovisual resources and reducing budgets of new productions. We focus on technologies specific to speech, text and image, respectively. These technologies incorporate efficient algorithms for processing and analysing relevant portions from various digital media and thus generating high-level semantic descriptors in the metadata space. After proposing an architecture for the integration of all the results of processing, we present indicative evaluation results in the context of CIMWOS.

Abstract: Rapid changes in the global marketplace have given rise to new demands and have provided new opportunities for the translation industry. The need for multilinguality in the presentation and business logic layers of most modern systems, applications and services is a great challenge that the translation industry now faces. But even after many years of intense research and many commercial attempts of related products, today's translation systems still fail to completely meet the above needs. Within this framework, an architecture of a modern automatic translation system exploiting current infrastructure and covering present and future needs is proposed in this paper.