Dimitra Anastasiou

Comparison of crowdsourcing translation with Machine Translation

In this paper we examine the model of crowdsourcing for translation and compare it with Machine Translation (MT). The large volume of material to be translated, the translation of this material into many languages combined with tight deadlines lead enterprises today to follow either crowdsourcing and/or MT. Crowdsourcing translation shares many characteristics with MT, as both can cope with high volume, perform at high speed, and reduce the translation cost. MT is an older technology, whereas crowdsourcing is a new phenomenon gaining much ground over time, mainly through Web 2.0. Examples and challenges of both models will be discussed and the paper is closed with future prospects regarding the combination of crowdsourcing and MT, so that they are not regarded as opponents. These prospects are partially based on the results of a survey we conducted. Based on our background, experience, and research, this paper covers aspects both from the point of view of translation studies and computational linguistics applications as well as of information sciences, and particularly the development of the Web regarding user-generated content.

Towards a Machine of a Process (MOP) ontology to facilitate e-commerce of industrial machinery

Adapting to users requirements is a key factor for enterprise success. Despite the existence of several approaches that point in this direction, simplifying integration and interoperability among users, suppliers and the enterprise during product lifecycle, is still an open issue. Ontologies have been used in some manufacturing applications and they promise to be a valid approach to model manufacturing resources of enterprises (e.g. machinery and raw material). Nevertheless, in this domain, most of the ontologies have been developed following methodologies based on development from scratch, thus ontologies previously developed have been discarded. Such ontological methodologies tend to hold the interoperability issues in some level. In this paper, a method that integrates ontology reuse with ontology validation and learning is presented. An upper (top-level) ontology for manufacturing was used as a reference to evaluate and to improve specific domain ontology. The evaluation procedure was based on the systemic methodology for ontology learning (SMOL). As a result of the application of SMOL, an ontology entitled Machine of a Process (MOP) was developed. The terminology included in MOP was validated by means of a text mining procedure called Term Frequency-Inverse Document Frequency (TF-IDF) which was carried out on documents from the domain in this study. Competency questions were performed on preexisting domain ontologies and MOP, proving that this new ontology has a performance better than the domain ontologies used as seed.

Gestures in assisted living environments

This paper is concerned with multimodal assisted living environments, particularly based on gesture interaction. The research of ambient assisted living is about the provision of a safe, comfortable, and inde-pendent lifestyle at a domestic environment. We refer to spatial gestures and gesture recognition software and present an observational user study related to gestures in the Bremen Ambient Assisted Living Lab (BAALL), a 60m2 apart-ment suitable for the elderly and people with physical or cognitive impairments.

A German-Chinese speech-gesture behavioural corpus of device control in a smart home

This paper reports on empirical studies on the use of speech and gesture interaction for device control in the Bremen Ambient Assisted Living Lab (BAALL) at the German Research Center for Artificial Intelligence at the University of Bremen. Our goal is to create and evaluate a multimodal corpus of device control (turn on lights, open sliding door, etc.). A Wizard-of-Oz user study was conducted in this lab with German and Chinese participants in order to carry out a comparative analysis of cross-lingual spatial spoken and gesture commands and report on cultural differences. The long-term goal is to design and develop a speech and gesture combining multimodal interactive system, while considering relevant behavioural aspects, such as cultural differences presented in this paper.

Gesture analysis in a case study with a tangible user interface for collaborative problem solving

This paper describes a case study that took place at the Public Research Centre Henri Tudor, Luxembourg in November 2012. A tangible user interface (TUI) was used in the context of collaborative problem solving. The task of participants was to explore the relation of external parameters on the production of electricity of a windmill presented on a tangible tabletop; these parameters were represented through physical objects. The goal of the study was to observe, analyze, and understand the interactions of multiple participants with the table while collaboratively solving a task. In this paper we focus on the gestures that the users performed during the experiment and the reaction of the other users to those gestures. Gestures were categorized into deictic/pointing, iconic, emblems, adaptors, and TUI-related. TUI-related/manipulative gestures, such as tracing and rotating, represented the biggest part, followed by the pointing gestures. In addition, we evaluated how active was the participation of the participants and whether gesture was accompanied by speech during the user study. Our case study can be described as a collaborative, problem solving, and cognitive activity, which showed that gesturing facilitates group focus, enhances collaboration among the participants, and encourages the use of epistemic actions.

Speech-to-speech translation in an assisted living lab

In this paper we describe speech processing in the Bremen ambient assisted living lab (BAALL); how it is currently and how it can be improved in the future with the implementation of Machine Translation (MT) technology. We are experiencing the ageing population phenomenon today and thus living in a safe, autonomous, independent, and convenient way in a domestic environment with multimodal human-machine interfaces is desired. We argue for multilinguality in assisted living environments by means of development of a speech-to-speech translation (SST) system.

COPSE: Rapidly Instantiating Problem Solving Activities based on Tangible Tabletop Interfaces

Collaborative problem solving is a skill that has become very important in our everyday lives and is constantly gaining attention in educational settings. In this paper, we present COPSE: a novel and unique software framework for instantiating Microworlds as collaborative problem solving activities on tangible tabletop interfaces. The framework provides three types of building blocks: widgets (provide input and localized feedback), equations (define the model), and scenes (visualize feedback), which can be specified in the form of structured text. Aim of COPSE is to simplify processes of creating, adjusting, and reusing custom Microworlds scenarios. We describe the structure of the framework, provide an example of a scenario, and report on a case study where we have used COPSE together with 33 teachers to build new scenarios on the fly.

Gestures used by intelligent wheelchair users

This paper is concerned with the modality of gestures in communication between an intelligent wheelchair and a human user. Gestures can enable and facilitate human-robot interaction (HRI) and go beyond familiar pointing gestures considering also context-related, subtle, implicit gestural and vocal instructions that can enable a service. Some findings of a user study related to gestures are presented in this paper; the study took place at the Bremen Ambient Assisted Living Lab, a 60m2 apartment suitable for the elderly and people with physical or cognitive impairments.

Case Study Analysis on Collaborative Problem Solving Using a Tangible Interface

In our research within the “Gestures in Tangible User Interfaces” project we examine how gestural and speech behaviour while solving a collaborative problem on tangible interfaces has an effect on 21st Century skills. This paper presents an explorative case study where 15–22 year old pupils of a public school in Luxembourg solved a collaborative problem using a tangible user interface (TUI). The main goal of the study was to investigate the frequency and kinds of gestures and the link between the usage of gestures with collaboration and problem solving. At the end of the study, participants had to fill in three standard post-test questionnaires about workload and usability, the results of which are presented here and show that the use of a TUI as an educational medium was regarded as a straightforward and simple solution that many people could learn to use very quickly. This promising result is in line with our research vision, which is to include TUIs in a collaborative setting educational environment.

A Questionnaire-based Case Study on Feedback by a Tangible Interface

This paper presents an intelligent tangible user interface used for learning in a collaborative setting in a secondary school environment. We describe our explorative case study with 48 participants and report the results of a questionnaire filled in after the end of the case study. The questionnaire is related to feedback modalities, such as visual, audio/aural, and haptic, which the participants would like to have offered in the future by the tangible interface or the tangible objects.