KM - Cognitive Computing overview by Ken Martin 13Apr2016
2 likes647 views
This document provides an introduction to cognitive computing and how it relates to knowledge management strategies. It begins with an overview of Ken Martin's background and the agenda. It then defines key cognitive computing concepts and technologies like natural language processing, machine learning, and pattern recognition. The document contrasts traditional and cognitive systems, noting cognitive systems are interactive, self-learning, and expand conversations. It maps cognitive capabilities to the KM lifecycle, showing how capabilities like natural language processing, text mining, and social network analysis can enhance each stage.
KM - Cognitive Computing overview by Ken Martin 13Apr2016
- 1. Cognitive Computing Discussion document An introduction to cognitive systems tools and techniques that influence KM practitioners development of strategies and solutions. Presenter: Ken Martin Email: Ken@KenMartinD.com
- 2. 2 Introduction to Ken Martin Ken Martin Summary work experience • 25+ Years delivering business and IT transformations to National and International organizations such as: Swiss Re, USAA, MetLife, Boeing; US Navy, US Airforce, NIH, VA, GSA, Microsoft Retail • Leadership and practice development: Oracle, KPMG, Bearing Point, Atos Origin, HCL • Developed web-based self-service Knowledge Analytics platform 2004-2008 – Innovator KM World Magazine: Knowledge Workers Marketplace Posted June 22, 2007 – Analytics kernel (eKM), which helps companies and their personnel transform information into reusable knowledge solutions. Currently Center of Excellence Director for Business Analytics Services at HCL • Strategic and advisory services in BI, Big Data, and Applied Analytics • Big Data and Advanced Analytics • Agile Analytics enterprise methodology delivery cross-functional business, data, and analytical services 19-Jan-2016
- 3. 3 Agenda 1 What you can expect from today’s presentation 2 Level setting on the basic terminology used to describe Cogitative Computing 4 Evolution from content creation and curation to applied analytical analysis and interaction 5 Venn Diagram of high-level Cognitive eco-system 6 Mapping Cognitive capabilities to the KM lifecycle 7 Contrasting Traditional and Cognitive Systems 8 Group discussion 19-Jan-2016
- 4. 4 What you can expect from today’s session • Introduction to the basics terminology and technologies involved in Cognitive Computing • Relating Cognitive Computing to the KM lifecycle with brief descriptions of key capabilities that enhance delivering KM solutions • Contrasting the current (traditional document management / content management, internet capabilities) with cognitive computing technologies. • Summarize the capabilities by mapping the KM lifecycle and cogitative computing capabilities • Benefits of using cognitive computing technologies to enhance KM solutions • Group discussion on customer interests and areas where cognitive computing impacts KM strategies 19-Jan-2016
- 5. Terminology Cognition Noun conscious mental activities : the activities of thinking, understanding, learning, and remembering Cognitive Adjective 1: of, relating to, being, or involving conscious intellectual activity (as thinking, reasoning, or remembering) 2: based on or capable of being reduced to empirical factual knowledge Cognitive computing is the simulation of human thought processes in a computerized model. Cognitive computing involves self-learning systems that use data mining, pattern recognition and natural language processing to mimic the way the human brain works. Pattern recognition is a branch of machine learning that focuses on the recognition of patterns and regularities in data, although it is in some cases considered to be nearly synonymous with machine learning. Machine learning explores the study and construction of algorithms that can learn from and make predictions on data. Classification Association Clustering Artificial intelligence Noun 1 : a branch of computer science dealing with the simulation of intelligent behavior in computers 2 : the capability of a machine to imitate intelligent human behavior • Dictionaries • Lexicons • Ontology • Categorizations • Natural Language 519-Jan-2016
- 6. Cognitive ComputingTraditional The evolution from content creation and curation to applied analytical analysis and interaction User interaction • Static Library • Taxonomy drive hierarchy • FAQ • Library of documents Applications • Static content • Hierarchical search • Ask and answer • Limited social connections • Document management • Relational database systems • Enterprise file management systems • Limited workflow automation Process & Store • Internet of Things • Any data / any device • Integration of sensors • Interactive interfaces support the KM lifecycle • From create to share and enhance of content • Continuous capture, analyze publish • Machine learning categorization • Advanced search –multidimensional • Contextual analysis proposes new questions • Shared social connections • Store any data, any time, any volume • Pattern recognition and machine learning analysis and recommendations • Integration workflow across communities 619-Jan-2016
- 7. Technologies behind Cognitive Computing Cognitive Computing: simplified a broad field of study, sensor technology, computer science, data science, behavioral science, neuro science, knowledge management and applied analytics. Sharing information • Written and Spoken language • Vision • Hearing Limited or does not exist in digital formats • Smell • Taste • Touch • Feel • Time • Location • Bio-markers / measurements • Voice recognition • Text to speech • Preferences • Images, videos • Language / translations • Time • Geo-special • Heat • Temperature • Pressure, depth • Velocity, altitude, distance • Chemicals • Vibration, light waves etc. 719-Jan-2016 Sensors Machine Learning Automation Content Curation Social Media Text Mining SemanticsPredictive Communications Prescriptive Recommend Cognitive Computing Corpus of data Warbles Internet of Things Devices measuring People sharing Social Networks Enterprise ApplicationsExternal Applications Advanced Analytics Computer Science Data Science Contextual learning & responding Vehicles Homes
- 8. Providing flexibility and agility to executing KM Strategies Observation: Applied and theoretical “Knowledge” capture within the human experience then transferred and retained digitally in books, manuals, videos, emails, spoken etc. Hypothesis: Using cognitive systems provides Knowledge Management and Strategy practitioners with effective tools throughout the KM lifecycle. KM Lifecycle stage Brief description of how cognitive system enhance KM Management and strategy Work / Experience Individuals and organizations digitally store work history, resume, time-on-job, white-papers, presentations, job locations etc. Create Any type of digital content: Text, video, photo, voice, geo-special, sensors Identify Automated search creates a metadata catalog: Authors, reviewers, devices, locations, accreditations Value Machine learning identifies: Work history, peer-reviews, accreditations Capture Big Data storage of any data: structured, semi-structured (sensor), and unstructured data Organize Machine learning recommendation of categorization, association, and clustering by topic, region, KM domain, security policies etc. Validate Social network analysis (Expertise, peer-reviews) Content analytics identifies usage, extracts terms, authors, feedback, and peer-reviews Document Advanced text mining and text analytics links sources and methods with authors and papers. (Disambiguation and Contextual correlation) Publish Crowd sourcing of documents, reviews, and methods is managed based on roles, responsibilities and security profiles Share Secure role-based distribution via advanced search connecting people and social networks based on security profile Apply Extractions and use captured as (time-on-page, downloads, contributions, reviews) survey cycles, mentions in emails and other references Measure Applied analytics provided via rich-interactive interface at the individual, social network group and organizational level Enhance Updates are monitored and routed near-real-time (accounting for reviews and approvals) based on changes in content and feedback Retire Outdated content is retained per the organization policy and mapped within the metadata catalog when retired 819-Jan-2016
- 9. Contrasting Traditional and Cognitive Systems Traditional system • Pre-configured answers and rules based recommendations • Decision support system providing results based past performance and standard math • Content and document management systems connected to search features Cognitive systems Automated interactive and self-learning systems connecting people to knowledge assets while expanding the conversation through the introduction of new content and insights. A cognitive system: 1. Learns from interactions based on a variety of inputs including text, speech, gestures, sensors (heat, pressure, time voltage, auditory levels), images, motion etc. 2. Adapts to language, timing, situational (cultural, life-stage, life-event) 3. Interacts and responds to the environment via (displays, text, speech, adjusting settings, movement, etc.) 4. Identifies context: Geo-special, cultural life (age)-stage, life-event , syntax, domain (knowledge and experience) 5. Extends context within limited information by proposing questions: Propose a question, respond to a question 6. Retains information and interactions based on measured outcomes 7. Retires knowledge assets when they are not longer used or content is outdated 919-Jan-2016
- 10. Wrap-up: Mapping the KM lifecycle to sources and Cognitive capabilities Enterprise data sources • Structure data (Employees, contractors) • Personal profiles • Education • Work experience External Sources • Other public data (by subject / domain) • Historical submission (matches) • Internal “Wiki” data • Social submissions • Research (shared libraries) Create Identify Value Capture Organize Validate Document Publish Share Apply Measure Enhance Retire Identify & Capture Evaluate & Enrich Publish & Apply Sustain Retire KM Officers and Sponsors Communities of Practice Exemplar sets / library • Notes and reports • Work papers (RFPs, Responses) • Standard of Practice • Operating Procedures Natural Language Processing Text Mining • Themes / summarize content • Categorize • Catalog Text Analytics • Link phrasing, subjects etc. Social Network Analysis • Link authors & reviewers • RACI Matrix • Security profiles Automated workflow • Interactive dashboards connect content to reviewers, approvers Cognitive interactions Machine learning and patter matching parses the document for relevance and generates hypothesis Evaluation is based KM structure and exemplar sets developed as part of the analysis process Reduces the time to (value) submission with automated analyze and categorize (informs ontology and taxonomy) document Categorization are viewed in rich visualization and tested for consistency and accuracy Internet of Things & Big Data System capture any data / any time / any volume Distribution and interactions to any device 1019-Jan-2016
- 11. Benefits to applying Cognitive Computing capability in your KM strategy • Extend expertise and personalized delivery of knowledge assets • Simplify the submission and curation of knowledge assets • Reduce the time to market for KM solutions • Promote adoption with rich content, simplified access, and personalized services • Provide additional enriched guidance as right-time responses, feedback, questions • Deliver highly valued KM assets, fastest path to the answer • Improve connections to the right people, their expertise, and associated knowledge assets 1119-Jan-2016
- 12. 12 Posing the questions for our group discussion Let’s have a group discussion 1. Are your customer using or considering cognitive systems? 2. What capabilities do you see being used by 1st adopters? 3. What barriers to adoption do you see across your customers? 4. How do you envision cognitive technology impacting f KM strategy and methods? 19-Jan-2016
- 13. 13 Thank you for participating 19-Jan-2016
- 14. Additional information Cognitive computing platforms • IBM Watson • Numenta • Google Now • Enterra Cognitive Reasoning Platform (CRP) • Saffron Emerging • MS Cortana • Open source custom niche’ system based on Big Data and Advanced analytics Use-cases include Across-industries • Help / service desk • Training • Predictive Maintenance Insurance / Financial • Customer Services • Fraud detection • Changes in customers behavior / life-event • Multiple Industries 14 It’s time to add KM to the list of growing investment in Cognitive Computing 19-Jan-2016
- 15. 15 Detailed Venn diagram of the Cognitive Computing ecosystem 19-Jan-2016