Towards Multi-modal Cognitive Big Data Informatics: Real World Case Studies & Future Challenges Professor AMIR HUSSAIN, PhD, Fellow UK HEA, Senior Member IEEE Founding Chief-Editor: Cognitive Computation journal (Springer, USA: http://springer.com/12559) Founding Chief-Editor: BMC Big Data Analytics journal (BioMed Central: http://bdataanalytics.com) Publications Chair, Annual 2015 INNS Big Data Conference (http://innsbigdata.org) Ass.Editor:IEEE Comp.Intelligence Mag.& IEEE Trans.Neural Networks & Learning Sys. PhD Director & Founding Director, Cognitive Big Data Informatics (CogBDI) Lab Division of Computing Science & Maths, University of Stirling, Scotland, UK E-mail: ahu@cs.stir.ac.uk http://cs.stir.ac.uk/~ahu/ http://cosipra.cs.stir.ac.uk (Journal published by Springer Neuroscience) Editor-in-Chief: Amir Hussain Computing Science Department has been ranked 10 th in UK in the recent UK Guardian Newspaper s University League Table (Computing Science & Informatics) 2011 Stirling University has been ranked in the top 50 in the world in 2015 The Times Higher Education 100 Under 50 table, which ranks world's best 100 universities under 50 years old Nearly two-thirds of our research has been rated Internationally Leading or Excellent in the most recent 2014 UK Research Excellence Framework (REF) assessment exercise (carried out by the UK government!) H2020 R&I in Europe, 5-6 Nov 2015
Overview of Scientific& Technological Excellence-I Cognitive Big Data Informatics: a trans-disciplinary research challenge of highest order! WHY? Understanding our own cognitive Big Data processing powers: how they are created and fostered & how they can go wrong due to brain malfunction; modelling of cognitive brain is an important step in developing such understanding Creating autonomous robots and agents able to think and act cognitively and ethically: Can support us in our daily lives in a variety of innovative ways AIMS: Carry out trans-disciplinary research into Cognitive Big Data Computation in order to: - Develop a comprehensive unified understanding of brain s Big Data processing capabilities perception, action, and attention; learning and memory; decision making and reasoning; language processing and communication; problem solving and consciousness; social cognition - Innovative and advance cognitive technology: Industry, commerce, robotics & many other areas are calling for creation of cognitive machines, with multi-modal cognitive Big Data analytic powers similar to those of ourselves, i.e. are able to think for themselves & reach decisions on actions in a variety of ways are flexible, adaptive & able to learn from their own previous experience & of others around them TARGET AREAS of Cognitive Big Data Research: Algorithmic, theoretical & computational approaches (e.g.online incremental &domain independent learning, nature & brain-inspired multi-modal cognitive learning,visualization &informatics); Implementations (e.g.neuromorphic, GPUs, smart clusters & clouds,open-source software); & Applications(e.g NLP, (e& m) healthcare, bio & neuro-informatics, natural robotics &space rovers, smart cities, social media & network analytics, surveillance, multimedia &business intelligence, power, energy &economic management)
Overview of Scientific& Technological Excellence-II Selected real-world Case Studies (led by CogBDI Lab -with 18 Postdoctoral & Doctoral researchers) Funded by over EUR2m grants from UK research councils, ESF, EU FP7 programs, UK & international charities & SME/industry, in collaboration with numerous UK, EU & international partners (including Harvard Medical School & MIT Media Lab) On-going research has led to 270+ high-impact research publications & commercialization of prototypes/patents For current/future Research Challenges in these &/or other new areas, new H2020 EU partners & collaborators very welcome! (numerous collaborating UK SME partners already on-board/committed) Case Study I:Multimodal,Social Big Data Sentiment &Opinion analytics (from Natural Language Text,Audio&Video) -realize emotional multi-lingual cognitive machines (including emotion-sensitive multi-modal interfaces) Case Study II:Multimodal informatics for Preventative(e&m)Healthcare, Smart Prosthesis/Assistive Technology (e.g.hearing-aids which can see,personalized &collaborative care, social e-companions for elderly & children) -realize natural interaction, visual & language processing capabilities in future SOCIO-COGNITIVE machines Case Study III:Multimodal autonomous cognitive (robotic, transportation & planetary rover) systems -realize motor action-selection& on-line learning capabilities in cognitive machines (e.g.in SMART SPACE CITIES)
Example Results from EU/International/UK RTD Projects Books(Selected from 16 to-date) [1] Cambria E, Hussain A Sentic Computing: A Common-Sense-Based Framework for Concept Level Sentiment Analysis, Springer Book Series on Socio-Affective Computing, in press, 2015 http://www.springer.com/series/13199 [2] Abel A, Hussain, Cognitively Inspired Audiovisual Speech Filtering: Towards an Intelligent, Fuzzy Based, Multimodal, Two-Stage Speech Enhancement System, SpringerBriefs in Cognitive Computation, Springer, (130 pages), 2015 High-impact Journal Papers (selected from 90+ to-date) [1] Hussain A, Cambria E, Schuller B (2014), Affective neural networks and cognitive learning systems for big data analysis, (Elsevier) Neural Networks, 58:1-3, 2014 (ISI-SCI Impact Factor: 2.7) [2] Poria S, Cambria E, Gelbukh A, Bisio F, Hussain A (2015), Sentiment Big Data Flow Analysis by Means of Dynamic Linguistic Patterns, IEEE Computational Intelligence Magazine, 10(4), 26-36, 2015 (ISI-SCI Impact Factor (IF): 2.7) [3] Malik, Z.K., Hussain, A. and Wu, Q.M.J. (2015), Multi-Layered Echo State Machine: A novel Architecture and Algorithm for Big Data applications, IEEE Transactions on Cybernetics, 2015 (accepted) (ISI-SCI IF: 3.469) Hussain A, Tao D, Wu J, Zhao D (2015), Computational Intelligence for Changing Environments, IEEE Comp. Intell. Mag., 10(4),10-11, 2015 [4] Malik, Z.K., Hussain, A., and Wu, J. (2015) An online generalized eigenvalue version of Laplacian Eigenmap for Visualing Big Data, in press, (Elsevier) Neurocomputing (ISI-SCI Impact Factor: 2), Sep 2015 (doi:10.1016/j.neucom.2014.12.119) [5] Poria S, Cambria E, Hussain A, Huang G, (2015) Towards an intelligent framework for multimodal affective data analysis, (Elsevier) Neural Networks (doi:10.1016/j.neunet.2014.10.005) 63, 104-116, March 2015 (ISI-SCI IF: 2.7) [6] Poria S, Cambria E, Hussain A, (2015) Fusing Audio, Visual and Textual Clues for Big Social Data Analysis, in press, Sep 2015 (Elsevier) Neurocomputing, (doi:10.1016/j.neucom.2014.12.119) (ISI-SCI Impact Factor: 2) [7] Poria S, Gelbukh A, Cambria, Hussain A, Huang G, (2014) Emo-SenticNet: Development and Applications, (Elsevier) Knowledge-Based Systems (KBS), 69:108 123, October 2014 (ISI-SCI Impact Factor: 4.1) [8] Poria, S. Gelbukh, A., Hussain, A., Das, D., Bandyopadhyay, S. (2013) Enhanced SenticNet with Affective Labels for Concept-based Opinion Mining, IEEE Intelligent Systems journal, 28(2), 31-38, March-April 2013 (ISI-SCI Impact Factor (IF): 3.1)
Forthcoming H2020 Calls of Interest& Partners Sought ICT-14: Big Data PPP: Cross-sectorial and cross-lingual data integration and experimentation (also know as "Innovation Spaces") ICT-15: Big Data PPP: Large Scale Pilot actions in sectors best benefitting from datadriven innovation (also known as "Lighthouse Projects") ICT-17: Big data PPP: Support, industrial skills, benchmarking and evaluation ICT-18: Big data PPP: Privacy-preserving big data technologies -Open to joining any existing consortium (with complementary interests/expertise, particularly real-world/industrial/benchmark problems/big data/applications providers & evaluators) -Also looking for new partners (SMEs, Universities and RTD institutions) for forming/leading new consortia in above or other Calls of mutual interest