from data silos to multi-sources and multi-agents cognitive platforms

Slides:



Advertisements
Similar presentations
Implementing Tableau Server in an Enterprise Environment
Advertisements

GMD German National Research Center for Information Technology Darmstadt University of Technology Perspectives and Priorities for Digital Libraries Research.
Introduction to Multimedia Adeyemi Adeniyi Bsc, MCP MCTS
Provenance in Open Distributed Information Systems Syed Imran Jami PhD Candidate FAST-NU.
ICT work programme ICT 22 Multimodal and natural computer interaction Aleksandra Wesolowska (Unit G.3 - Data Value Chain) Juan Pelegrin (Unit.
Presented by Mina Haratiannezhadi 1.  publishing, editing and modifying content  maintenance  central interface  manage workflows 2.
1 INTRODUCTION TO DATABASE MANAGEMENT SYSTEM L E C T U R E
Artificial Intelligence By Michelle Witcofsky And Evan Flanagan.
NETWORK HARDWARE AND SOFTWARE MR ROSS UNIT 3 IT APPLICATIONS.
WEB MINING. In recent years the growth of the World Wide Web exceeded all expectations. Today there are several billions of HTML documents, pictures and.
CS 127 Introduction to Computer Science. What is a computer?  “A machine that stores and manipulates information under the control of a changeable program”
ICT-enabled Agricultural Science for Development Scenarios, Opportunities, Issues by ICTs transforming agricultural science, research & technology generation.
Digital Video Library Network Supervisor: Prof. Michael Lyu Student: Ma Chak Kei, Jacky.
Big traffic data processing framework for intelligent monitoring and recording systems 學生 : 賴弘偉 教授 : 許毅然 作者 : Yingjie Xia a, JinlongChen a,b,n, XindaiLu.
Authoring tools There are three main authoring tools:
Introduction to Multimedia. What is Multimedia? Derived from the word “Multi” and “Media” Multi Many, Multiple, Media Tools that is used to represent.
CS 784: Advanced Topics in Data Management This semester’s focus: Data Science AnHai Doan.
Dato Confidential 1 Danny Bickson Co-Founder. Dato Confidential 2 Successful apps in 2015 must be intelligent Machine learning key to next-gen apps Recommenders.
Glencoe Introduction to Multimedia Chapter 2 Multimedia Online 1 Internet A huge network that connects computers all over the world. Show Definition.
INHA UNIVERSITY, KOREA Rainer Simon Austrian Institute of Technology.
Book web site:
ecommerce + affiliate marketing 2003 ecommerce store 2004 – 2006 ecommerce + affiliate marketing 2007 – present TopLine Media Group.
The World Wide Web.
Multi-Device UI Development for Task-Continuous Cross-Channel Web Applications Enes Yigitbas, Thomas Kern, Patrick Urban, Stefan Sauer
4/19/ :02 AM © Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN.
Containers as a Service with Docker to Extend an Open Platform
HTTP AND ABSTRACTION ON THE INTERNET
TC 310 The Computer in Technical Communication
Big Data A Quick Review on Analytical Tools
Unit 9 Writing for New Media
Intelligent Agent Solution
Bridges and Clouds Sergiu Sanielevici, PSC Director of User Support for Scientific Applications October 12, 2017 © 2017 Pittsburgh Supercomputing Center.
Introduction to Azure Bot Framework
Extensible Platform Microsoft Dynamics 365
Unit - 5 Activity - 1 Competency – 5
CHAPTER 8 Multimedia Authoring Tools
Microsoft Build /13/2018 2:24 AM © 2016 Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY,
Introduction to J2EE Architecture
Web 3.0 and its Impact on E-Business
Intelligent Agents Chapter 2.
H3 Solutions and the Azure Government Cloud Team Up to Power Contextual Intelligence Platform – Where Big Data Meets Business Productivity MICROSOFT AZURE.
Data Representation – Instructions
CIS 515 STUDY Lessons in Excellence-- cis515study.com.
Information Systems in Organizations 1.1 Introduction to MIS
Big Data The huge amount of data being collected and stored about individuals, items, and activities and to the process of drawing useful information from.
Information Systems in Organizations 1.1 Introduction to MIS
Media, Multimedia & Digital Media
Information Systems in Organizations 1.1 Introduction to MIS
Accelerate application delivery with a Cloud-native mindset
Foundations of Networking
1 مفهوم ارتباطات ارتباطات معادل واژه communications ) ميباشد(. ارتباطات يك فرايند اجتماعي و دو طرفه است كه در آن اطلاعات مبادله شده و نوعي تفاهم بين طرفهاي.
Microsoft Ignite NZ October 2016 SKYCITY, Auckland.
Foundations of Networking
Introduction to Multimedia
Internet Protocols IP: Internet Protocol
Requirements Management
Automating Profitable Growth™
Dynamicweb PIM General introduction Innovia 2018.
Intro to html5.
Information Systems in Organizations 1.1 Introduction to MIS
LO3 – Understand Business IT Systems
FashionBrain: Understanding Europe’s Fashion Data Universe
Educational Computing
Chaitali Gupta, Madhusudhan Govindaraju
Web archives as a research subject
Information Systems in Organizations 1.1 Introduction to MIS
A DevOps process for deploying R to production
Bridge the Gap Between Statistician and Data Analysis Professionals
Exploring Cognitive Services
R for Data Science Data science Data science is a booming field in today’s world. Since Artificial Intelligence is the main focus of today’s technology,
Presentation transcript:

from data silos to multi-sources and multi-agents cognitive platforms Aymeric, Brisse, CTO of Perfect Memory. So today I am going to make a short presentation about multi-agents cognitive platforms. What is it, and for what type of scenarios it is made for. AYMERIC BRISSE XLDB 2017 – LIGHTNING TALK

Big Data | Applications | New challenges NoSQL & distributed FS address the storage and access issues (Cassandra, CephFS, etc.) Data scientist & Machine Learning address the valuable information extraction issues This allows awesome services to be built (image recognition, recommandation, etc.) How to make these different services & data silos interoperable to cross-reference the data? How to make intelligible the mass of data now available to meet specific business needs? | New challenges To resolve the main issues brought by Big Data, we had to invent new tools. To be able to store and to access so much data, we have built new databases & new filesystems like Cassandra & CephFS To be able to find in that tools valuable information, a new computer science field has emerged, the data science with tools like the ML We have seen in the last years a lot of awesome new services like image recognition and recommendation used by thousand of companies. And that brings new challenges too.

Use case | Example A TV Channel receives various content (footages, pictures, wires) from dozens of news agencies Automate the valuable information extraction Retrieve structured & intelligible information (who, where, what) Let’s take a simple use case to illustrate these challenges. A TV channel receives binary content like footages from dozens of news agency. And that TV channel wants to automate the information extraction. By information we talk about structured & intelligible data). For that, they want to use a cognitive platform

Cognitive Platform | Workflow driven, based on multi-agents Basically, a cognitive platform is able to adapt the workflow to the content it process, it will know what to do and how, and will use for that use multiple agents. Each agent is dedicated to a very specific task like Image Analysis. The platform brings a framework allowing each agent to act as a bridge between the service it uses and the platform semantic language. For example, a part of the workflow is dedicated to the audio track processing. So it will use first use for example an agent that is specialized in STT, while another agent will process that text to extract the named entities (persons, organizations, places) and finally another agent will retrieve additional information on the Linked Open Data (like Wikidata or dbpedia) on these entities. As the workflow is run, the knowledge graph of the footage is gradually built via the merging of each individual agent knowledge. So it’s really the cooperation made available by the cognitive platform between them allows high level questions to be answered: what’s going on in that video?

Knowledge Graph | Exploitable & Interoperable information Because in the end the TV channel wants to obtain structured information like : This physics-related video displays the Physicist Rainer Weiss and talks about the gravitational waves.

Stack | Under the hood Base - Openshift (Kubernetes with Docker containers & DevOps tools) Runtime - One Docker image per Agent Communication - RabbitMQ for message delivery Databases - Various Storage - Ceph DevOps Philosophy - « 12-Factor App » (https://12factor.net/) @ Perfect Memory has developed its cognitive platform, and you can find on that page some information on the stack we use. So the platform runs on Openshift. It means that each agent runs in it own docker container, and communicates with other agents with RabbitMQ For knowledge access & storage, we use various databases, mainly ontotext for the semantic graph database. For files storage we use the distributed FS ceph with rados And last but not least, we did implement all the best practices of the “12 factor app” manifest which helped us a lot 6

Aymeric BRISSE - CTO aymeric.brisse@perfect-memory.com OK that’s it, I will be happy to give you more details on the platform after during the poster session Aymeric BRISSE - CTO aymeric.brisse@perfect-memory.com