![]() |
ISCApad #179 |
Friday, May 10, 2013 by Chris Wellekens |
7-1 | ACM TiiS special issue on Machine Learning for Multiple Modalities in Interactive Systems and Robots Call for Papers Special Issue of the ACM Transactions on Interactive Intelligent Systems on MACHINE LEARNING FOR MULTIPLE MODALITIES IN INTERACTIVE SYSTEMS AND ROBOTS Main submission deadline: February 28th, 2013 http://tiis.acm.org/special-issues.html AIMS AND SCOPE This special issue will highlight research that applies machine learning to robots and other systems that interact with users through more than one modality, such as speech, touch, gestures, and vision. Interactive systems such as multimodal interfaces, robots, and virtual agents often use some combination of these modalities to communicate meaningfully. For example, a robot may coordinate its speech with its actions, taking into account visual feedback during their execution. Alternatively, a multimodal system can adapt its input and output modalities to the user's goals, workload, and surroundings. Machine learning provides interactive systems with opportunities to improve performance not only of individual components but also of the system as a whole. However, machine learning methods that encompass multiple modalities of an interactive system are still relatively hard to find. This special issue aims to help fill this gap. The dimensions listed below indicate the range of work that is relevant to the special issue. Each article will normally represent one or more points on each of these dimensions. In case of doubt about the relevance of your topic, please contact the special issue associate editors. TOPIC DIMENSIONS System Types - Interactive robots - Embodied virtual characters - Avatars - Multimodal systems Machine Learning Paradigms - Reinforcement learning - Active learning - Supervised learning - Unsupervised learning - Any other learning paradigm Functions to Which Machine Learning Is Applied - Multimodal recognition and understanding in dialog with users - Multimodal generation to present information through several channels - Alignment of gestures with verbal output during interaction - Adaptation of system skills through interaction with human users - Any other functions, especially combining two or all of speech, touch, gestures, and vision SPECIAL ISSUE ASSOCIATE EDITORS - Heriberto Cuayahuitl, Heriot-Watt University, UK (contact: h.cuayahuitl[at]gmail[dot]com) - Lutz Frommberger, University of Bremen, Germany - Nina Dethlefs, Heriot-Watt University, UK - Antoine Raux, Honda Research Institute, USA - Matthew Marge, Carnegie Mellon University, USA - Hendrik Zender, Nuance Communications, Germany IMPORTANT DATES - By February 28th, 2013: Submission of manuscripts - By June 12th, 2013: Notification about decisions on initial submissions - By September 10th, 2013: Submission of revised manuscripts - By November 9th, 2013: Notification about decisions on revised manuscripts - By December 9th, 2013: Submission of manuscripts with final minor changes - Starting January, 2014: Publication of the special issue on the TiiS website, in the ACM Digital Library, and subsequently as a printed issue HOW TO SUBMIT Please see the instructions for authors on the TiiS website (tiis.acm.org). ABOUT ACM TiiS TiiS (pronounced 'T double-eye S'), launched in 2010, is an ACM journal for research about intelligent systems that people interact with.
| |||||
7-2 | CfP EURASIP Journal Special Issue on Informed Acoustic Source Separation CALL FOR PAPERS
| |||||
7-3 | EURASIP Journal on Advances in Signal Processing:*Special Issue on Informed Acoustic Source Separation* CALL FOR PAPERS
| |||||
7-4 | CfP International Journal of Computational Linguistics and Chinese Language Processing: Special Issue on Processing Lexical Tones in Natural Speech
Call for Papers International Journal of Computational Linguistics and Chinese Language Processing Special Issue on Processing Lexical Tones in Natural Speech
This special issue aims to address questions about how lexical tones are processed by humans and machines in the context of natural, continuous speech. Lexical tones in tone languages have been widely investigated in the fields of linguistics, psycholinguistics, computational linguistics, and language acquisition by applying a wide range of theoretical, empirical, and experimental approaches. As the phonetic representation of lexical tones which are produced in connected speech can differ considerably from that of lexical tones which are produced in isolation, research interests constantly grow in how lexical tones are produced, perceived, and processed in realistic speech data. This special issue aims to bring together methodologies from different research disciplines to extend our understanding of lexical tones used in real speaking situations. We welcome submissions addressing the following issues.
Paper submission deadline: J Notification of acceptance: May 31, 2013 Final paper due: August 31, 2013 Tentative publication date: December, 2013 All submitted papers should present original research work, which has not been published elsewhere. Submitted manuscripts will be peer-reviewed by at least two independent reviewers. For detailed submission guidelines, please visit the website of the International Journal of Computational Linguistics and Chinese Language Processingat http://www.aclclp.org.tw/journal/submit.php. Please also feel free to contact the Guest Editor of this special issue, Dr. Shu-Chuan Tseng, at tsengsc@gate.sinica.edu.tw, if you need any additional information.
| |||||
7-5 | CfP Numéro spécial de la revue TAL sur 'les Entités Nommées et leurs relations'PREMIER APPEL À CONTRIBUTIONS
Numéro spécial de la revue TAL sur 'les Entités Nommées et leurs relations'
Direction : Sophia Ananiadou (Sophia.ananiadou@manchester.ac.uk) Nathalie Friburger (nathalie.friburger@univ-tours.fr) Sophie Rosset (sophie.rosset@limsi.fr) Les entités nommées constituent un champ de recherche très actif depuis de nombreuses années. Elles sont depuis longtemps considérées comme un point central dans de nombreuses applications mettant en jeu des notions comme la compréhension, la recherche sémantique, etc. La notion d'Entités Nommées (EN) couvre non seulement les noms propres mais aussi des entités plus complexes comme les expressions multi-mots. Les entités nommées sont en général typées selon des taxonomies plus ou moins vastes et fortement dépendantes du domaine d'application ou des besoins considérés. Elles recouvrent classiquement des noms désignant des personnes, des lieux ou des organisations mais peuvent aussi se rapporter à des notions plus techniques comme les maladies. La détection des entités et de leurs relations, malgré de nombreuses années de recherche, reste un problème difficile. Les principaux points à résoudre incluent la résolution des ambiguïtés, la détection des synonymes, la co-référence et la variabilité (acronymes, orthographe etc.). Plusieurs méthodes ont été proposées et évaluées pour améliorer la détection et la classification des entités ainsi que leurs relations. Ces méthodes vont des approches fondées sur des connaissances explicites (approches à base de règles, de lexiques, etc.) aux approches à base d'apprentissage supervisé, faiblement supervisé voire non supervisé. L'évaluation des systèmes de détection des entités nécessite au minimum l'existence de corpus de référence (gold standard). L'évaluation des relations ajoute un niveau de complexité puisqu'il s'agit, le plus souvent, d'une tâche complexe impliquant la détection des entités puis des relations entre elles. Comment évaluer la détection des relations dans le cadre d'une tâche complète ? Quelle métrique utiliser pour tenir compte des erreurs de la tâche précédente (détection des EN) ? Si les entités nommées simples permettent d'atteindre de bons voire de très bons résultats, il n'en va pas de même lorsque leur définition est complexe ou lorsqu'on traite des domaines spécialisés. Nous invitons donc les contributions portant sur tout aspect relatif au traitement des entités nommées, et en particulier (liste non exclusive) : - définition et typologie des entités nommées, y compris dans un sens étendu - détection des entités nommées et type de documents (résumés, articles, ressources collaboratives comme Wikipedia, domaine de spécialité, media sociaux comme twitter, emails, file de discussion, parole...) - détection des empans et analyse structurelle des EN - co-référence inter-documents et suivi d'entités - suivi d'entités à travers le temps, les groupes sociaux et géographiques, suivi d'entités intra- et inter-documents, etc. - reconnaissance d'entités nommées dans le domaine général ou en domaine de spécialité - guides et schémas d'annotation, outils, méthodes et corpus annotés - aspects multilingues, extraction d'entités dans des corpus comparables ou parallèles - désambiguïsation des entités - applications et entités - évaluation, comparaison et validation d'outils LANGUE Les articles sont écrits en français ou en anglais. Les soumissions en anglais ne sont acceptées que pour les auteurs non francophones. LA REVUE Depuis 40 ans, TAL (Traitement Automatique des Langues) est un journal international publié par l'ATALA (Association pour le Traitement Automatique des Langues) avec le soutien du CNRS. Depuis quelques années, il s'agit d'un journal en ligne, des versions papier pouvant être obtenues sur commande. Ceci n'affecte en rien le processus de relecture et de sélection. DATES IMPORTANTES * 15 avril 2013 Deadline pour la soumission * Juillet 2013 Notification aux auteurs * Automne 2013 Publication FORMAT DE LA SOUMISSION Les articles soumis doivent décrire un travail original, complet et non publié. Chaque soumission sera étudiée par 2 membres du comité de programme. Les articles (25 pages environ, format PDF) doivent être déposés sur la plateforme Sciencesconf [adresse disponible très bientôt] Merci de contacter les éditeurs de cette revue si vous souhaitez y soumettre un article en leur fournissant un résumé d'une page. Sophia Ananiadou (Sophia.ananiadou@manchester.ac.uk) Nathalie Friburger (nathalie.friburger@univ-tours.fr) Sophie Rosset (sophie.rosset@limsi.fr) Les papiers acceptés feront au maximum 25 pages en PDF. Le style est disponible pour téléchargement sur le site du journal (http://www.atala.org/-Revue-TAL-) COMITE SCIENTIFIQUE (tentative) Maud Ehrmann, European Commission, JRC Olivier Galibert, LNE, France Natalia Grabar, STL, Université de Lille 1 et 3, France Kais Haddar, University of Sfax, Tunisie Thierry Hamon, LIM&Bio, Paris 13, France Sanda Harabagiu, Texas, USA Valia Kordoni, Humboldt-Universität, Berlin, Germany Anna Korhonen, University of Cambridge, UK Ioannis Korkontzelos, University of Manchester, UK Anne-Laure Ligozat, LIMSI, France Bernardo Magnini, FBK, HLT, Italy Makoto Miwa, University of Manchester, UK Claire Nedellec, MIG, INRA, France Aurélie Névéol, LIMSI, France Noaoaki Okazaki, Tohoku University, Japan Christian Raymond, IRISA, France Fabio Rinaldi, University of Zurich Patrick Ruch, University of Geneva, Swiss Benoit Sagot, ALPAGE, France Satoshi Sekine, NYU, USA Jian Su, A-STAR, Singapore Junichi Tsujii, Microsoft Research Asia, China Patrick Watrin, UCL, CENTAL, Belgique Fabio Zanzotto, Tor Vergata, university of Rome, Italy
--
| |||||
7-6 | Special Issue of COMPUTER SPEECH AND LANGUAGE on Next Generation ParalinguisticsCall for Papers Special Issue of COMPUTER SPEECH AND LANGUAGE on Next Generation
Paralinguistics __________________________________
http://www.journals.elsevier.com/computer-speech-and-language/call-for-papers/next-generation-computational-paralinguistics/ Computational Paralinguistics recently reached a level of maturity allowing for their first real-life applications in interaction, coaching, media retrieval, robotics, surveillance, and manifold further
domains. In particular, an increasing level of realism is recently faced by coping with speaker
independent analysis of highly naturalistic data in narrow-bandwidth, noisy, or reverberated
conditions. At the same time, the richness of the range of speaker states and traits analysed
computationally is increasingly widening up. This includes in particular also the degree of
subjectivity faced with tasks such as perceived speaker personality, likability, or intelligibility,
to name a few. Both these aspects require additional experience on the interplay of states
and traits in speech, singing, and language. Further, with the integration in applications,
novel aspects arise such as efficiency, reliability, self-learning, mobility, multi-cultural and
multi-lingual aspects, handling groups of speaker or singers, standardisation, and user
experience with such systems. This Special Issue thus aims at shaping the Next Generation Computational Paralinguistics.
It will focus on technical issues for highly improved and reliable state and trait analysis in spoke
this topic. Original, previously unpublished submissions are encouraged within the following n,
sung, and written language and provide a forum for some of the very best experimental work
scope: +Analysis of States and Traits in Spoken, Sung, and Written Language +Subjectivity in Computational Paralinguistics (e.g., perceived states and traits) +Interdependence of States and Traits +Intelligibility of Language Varieties and Deviant Speech +Efficiency (low energy and memory consumption, fast adaptation, active learning, etc.) +Reliability (e.g., confidence measures, robustness against regulation and feigning, overlap) +Self-learning (unsupervised, partially supervised, reinforced, and deep learning) +Mobility (client/server distribution, package loss, coding artefacts, privacy preservation, etc.) +Multicultural and Multilingual Issues +Speaker / Singer Group Characterisation +Standardisation (output encoding, feature encoding, etc.) +Application (interaction, voice and writing coaching, retrieval, robotics, surveillance, etc.) +User Experience of Computational Paralinguistics Systems Important Dates __________________________________ Submission Deadline 1 April 2013 First Notification 1 July 2013 Final Version of Manuscripts 1 November 2013 Tentative Publication Date January 2014 Guest Editors __________________________________ Björn Schuller, Technische Universität München, Germany, schuller@IEEE.org Stefan Steidl, FAU, Germany, stefan.steidl@fau.de Anton Batliner, Technische Universität München, Germany, Anton.Batliner@lrz.uni-muenchen.de Alessandro Vinciarelli, University of Glasgow / IDIAP, UK, alessandro.vinciarelli@glasgow.ac.uk Felix Burkhardt, Deutsche Telekom AG, Germany, Felix.Burkhardt@telekom.de Rob van Son, University of Amsterdam / Netherlands Cancer Institute, NL , r.v.son@nki.nl Submission Procedure __________________________________ Prospective authors should follow the regular guidelines of the Computer Speech and
Language Journal for electronic submission (http://ees.elsevier.com/csl). During submission
authors must select for this Special Issue (short name 'NextGen Paralinguistics').
| |||||
7-7 | Special Issue of Neural Networks (Elsevier) on Affective and Cognitive Learning Systems for Big Social Data Analysis Special Issue of Neural Networks (Elsevier) on
Affective and Cognitive Learning Systems for Big Social Data Analysis
Guest Editors Amir Hussain*, University of Stirling, United Kingdom (ahu@cs.stir.ac.uk) Background and Motivation As the Web rapidly evolves, Web users are evolving with it. In an era of social connectedness, people are becoming more and more enthusiastic about interacting, sharing, and collaborating through social networks, online communities, blogs, Wikis, and other online collaborative media. In recent years, this collective intelligence has spread to many different areas, with particular focus on fields related to everyday life such as commerce, tourism, education, and health, causing the size of the Web to expand exponentially. The distillation of knowledge from such a large amount of unstructured information, however, is an extremely difficult task, as the contents of today's Web are perfectly suitable for human consumption, but remain hardly accessible to machines. The opportunity to capture the opinions of the general public about social events, political movements, company strategies, marketing campaigns, and product preferences has raised growing interest both within the scientific community, leading to many exciting open challenges, as well as in the business world, due to the remarkable benefits to be had from marketing and financial market prediction. Existing approaches to opinion mininig mainly rely on parts of text in which sentiment is explicitly expressed, e.g., through polarity terms or affect words (and their co-occurrence frequencies). However, opinions and sentiments are often conveyed implicitly through latent semantics, which make purely syntactical approaches ineffective. In this light, this Special Issue focuses on the introduction, presentation, and discussion of novel techniques that further develop and apply big data analysis tools and techniques for sentiment analysis. A key motivation for this Special Issue, in particular, is to explore the adoption of novel affective and cognitive learning systems to go beyond a mere word-level analysis of natural language text and provide novel concept-level tools and techniques that allow a more efficient passage from (unstructured) natural language to (structured) machine-processable data, in potentially any domain. Articles are thus invited in areas such as machine learning, weakly supervised learning, active learning, transfer learning, deep neural networks, novel neural and cognitive models, data mining, pattern recognition, knowledge-based systems, information retrieval, natural language processing, and big data computing. Topics include, but are not limited to: • Machine learning for big social data analysis The Special Issue also welcomes papers on specific application domains of big social data Timeframe Call for Papers out: April 2013 Composition and Review Procedures The Elsevier Neural Networks Special Issue on Affective and Cognitive Learning Systems for Big Social Data Analysis will consist of papers on novel methods and techniques that further develop and apply big data analysis tools and techniques in the context of opinion mining and sentiment analysis. Some papers may survey various aspects of the topic. The balance between these will be adjusted to maximize the issue's impact. All articles are expected to successfully negotiate the standard review procedures for Elsevier Neural Networks.
___________________________________________
Univ.-Prof. Dr.-Ing. habil. Björn W. Schuller
Head Institute for Sensor Systems University of Passau Passau / Germany
Head Machine Intelligence & Signal Processing Group Institute for Human-Machine Communication Technische Universität München Munich / Germany
CEO audEERING UG (haftungsbeschränkt) Gilching / Germany
Visiting Professor School of Computer Science and Technology Harbin Institute of Technology Harbin / P.R. China
Associate Institute for Information and Communication Technologies JOANNEUM RESEARCH Graz / Austria
Associate Centre Interfacultaire en Sciences Affectives Université de Genève Geneva / Switzerland
schuller@ieee.org http://www.schuller.it ___________________________________________
|