Cognitive Computing and Conversational Agents Rob High, Jr. IBM Fellow, Vice President Chief Technology Officer IBM Watson 1
Memex: Vannevar Bush, As We May Think, The Atlantic Monthly, 1945 A device for people to compress and store all of their books, records, and communications, mechanized to provide rapid recall, access and cross document linkages 2
1970 In from 3 to 8 years we will have a machine with the general intelligence of an average human being Marvin Minsky 1977 There is no reason anyone would want a computer in their home Ken Olson, Chairman DEC 3
DeepQA: Incremental Progress in Answering Precision on the Jeopardy Challenge: 6/2007-11/2010 100% Precision 90% 80% 70% 60% 50% 40% 30% 20% v0.6 10/09 v0.1 12/07 v0.5 05/09 v0.4 12/08 v0.3 08/08 IBM Watson Playing in the Winners Cloud v0.8 11/10 v0.2 05/08 V0.7 04/10 10% Baseline 12/06 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% % Answered 4
With Precision, Accurate Confidence and Speed, the rest was History 5
Why are Cognitive Systems Important to Modern Life? Digital information growth is exponential Understand the intention of human expression Act as a research assistant to pinpoint relevant information Amplify human abilities to scale the democratization of expertise 2.5 Exabytes of data produced every day This is where we were in 2016 This is where we will be in 2020 6
a By 2018 half of all consumers will regularly interact with services based on cognitive - IDC FutureScape 7
Cognitive systems amplify human cognition 8
What is Cognitive Computing? Cognitive systems understand human expressions textual, verbal, visual By reasoning about the actual intention or problem being addressed They learn how to recognize patterns of meaning through examples and feedback And they interact with humans on their own terms, and in a way that inspires people. and do so at enormous scale. 9
Some common concepts Analytic Systemsuse statistical techniques for detecting patterns within data Natural Language Processing (NLP)is a core function for parsing and identifying significant words in language Machine Learningis a class of statistical techniques that use training data to recognize the correlation between a set of feature patterns and outcomes Cognitive Systemsleverage machine learning to predict meaning in features of human language (spoken, written, visual) and related forms of human reasoning Deep Learningis a rapidly maturing space, based on neural network techniques, that are taught to find their own features Cognitive Solutionsleverage a combination of cognitive system reasoning strategies and other analytic and classical computing techniques to solve for a problem Yield an understanding of historical or current state from which to draw conclusions Can be used recognize trends, or predict unknown information Uses Annotators to identify significant words Increasingly used to extract Entities and Relationships into Knowledge Graphs Often rule-based, but a growing focus on statistical techniques Enables the prediction of other similar outcomes given a similar pattern of feature values Often associated with the problem of identifying the weak signals in the noise that indicate meaningful patterns Reasoning strategies often employ high-dimensional feature spaces A lot of work goes in to identifying and refining appropriate features Generally very high dimensional, fed with an arbitrarily high number of primitive signals and left to sort itself out through training data and reinforcement Learn from training data and experience; Interact to the human condition; Transparent about their rational and confidence; Evolve to improve Amplify Human Cognition 10
The Watson Cognitive Computing Evolution 2011: Jeopardy introduced a Factoid Pipeline for use with general domain knowledge 2012: Watson Discovery Advisor leverages the Factoid Pipeline around specific domains to help find the questions you re not thinking to ask Healthcare Utilization Management, Oncology Treatment Advisor, Clinical Trial Manager 2013: Watson Engagement Advisor Introduced a Passage and then FAQ pipeline 2014: Expanded Watson Discover Advisor added knowledge extraction and graph visualization Introduced Watson Explorer combined with Watson Content Analytics Watson Ecosystem made the (WEA) Passage and FAQ pipeline available as a service 2015: Watson Developer Cloud dramatically expanded the range of services designed to interpret the human condition, and made them available on Bluemix 2016: Enhancing Human Engagement focusing on conversational agents, with emotion detection and expression 2017: Developer experiences Cloud-native scaling and developer efficiency 11
Watson Reference Model 12
Watson is available as a set of services delivered as APIs in the Cloud ibm.com/bluemix Higher Reasoning Skills Higher Reasoning Skills Conversation Discovery Can be combined with the 100s of other available services on Bluemix 13
Conversation Add a natural language interface to your application to automate interactions with your end users. Common applications include virtual agents and chat bots that can integrate and communicate on any messaging platform. Model is trained on user-defined intents, entities and dialogs. Expanded to recognize the emotion of the user and to respond accordingly. 14
A cognitive chat-bot is a conversation A conversation should: Engage the user Focus on the user s broader concern Build on an idea Leave the user inspired and satisfied 15
A cognitive conversation can occur through many different channels Mobile Apps, Web Applications and IVR telephone calls Messaging Platforms Proprietary APIs Bot APIs Watson Conversation Facebook/M, Twitter, Twillio, Slack, etc. Watson Bot Framework Business Lo gic e.g. Reserve Table Personal Assistance Devices Amazon Echo, Google Home, Siri, etc. Even, in a collaboration multiple people in a conference room 16
Now, imagine you re in
Now, imagine you re in 18
Watson Virtual Agent WVA is a collection of 105 common, pre-built Conversations Account Management Billing Device Management Help / Information Complaints Order Management Payment Sales Service Management Just configure, add your App, and Go! 19
Discovery Service Entry point into deep reasoning over literary and structured resources Entities News Sentiment Time-sequenced Co-mentions and Trends Related concepts Over 300,000 news articles and blogs added daily Connect your enterprise data sources 20
Watson is available as a set of services delivered as APIs in the Cloud ibm.com/bluemix Knowledge Organization Skills Knowledge Organization Skills Watson Knowledge Studio Document Conversion Retrieve and Rank Can be combined with the 100s of other available services on Bluemix 21
Watson Knowledge Studio Enable subject matter experts and developers to teach Watson the linguistic nuances of industries and knowledge domains 22
Retrieve and Rank Uses machine learning to improve search across documents using natural language questions 23
Watson is available as a set of services delivered as APIs in the Cloud Watson Foundational Skills are grouped into four categories ibm.com/bluemix Foundational Cognitive Skills Language Author Concepts Dates Entities Feeds Keywords Language Microformats Publication Date Relations Typed Relations Sentiment Taxonomy Text Extraction Speech Speech to Text Telephony Speech to Text Keyword Spotting Vision Image Classification Face Detection and Attribution Celebrity Recognition Text to Speech Expressive Text to Speech Visual Text Recognition Similarity Searching Empathy Personality Insights Tone Analyzer Emotion Analysis Can be combined with the 100s of other available services on Bluemix 24
Expressive Text to Speech Enables speech that reflects intended tone Expressive SSML Voice Transformation SSML 25
Watson Visual Recognition Service Image Classification Face Detection and Attribution Celebrity Recognition Visual Text Recognition Similarity Searching 26
Tone Analyzer Understands and Helps Fine Tune Your Message Uses psycholinguistics, emotion analysis and language analysis to assess Tone I'm a hard working adventurous, very talented man who's been caring and helpful throughout my life, I like to travel, play my guitar, dance, and cook, I love the beach, sailing my boat, and the outdoors. I raised two great kids and now I'm starting a new chapter in my life. Thanks. What I m doing with my life Working toward a new goal, keeping fit, helping others, and traveling whenever I get a chance. I m really good at Listening, enjoying the moment, and many other things. The six things I could never do without Family, the ocean, intimacy, friends, adventure, music, love. On a typical Friday night I am Meeting with friends, listening to a band or playing my guitar, dancing or just staying home with someone special and enjoying each other. You should message me if You're looking for a relationship with someone that likes to sail his boat, ride bicycles, travel, swim, go to the beach, listen to music and enjoy everyday pleasures together. 27
You also need to know how people are feeling
You also need to know how people are feeling 29
Emotional Analysis helps build empathetic systems Uses state-of-the-art machine learning models and feature engineering techniques to predict emotion labels 30
Personality Insights 31
Many institutions are already building advantage from conversational chat-bots H&R Block Tax preparation Staples/Easy System Buy on-demand Autodesk/Otto Problem resolution The Weather Channel Personalized weather Royal Bank of Scotland Account management 1-800-Flowers/Gwyn Gift concierge Meeka Wedding Planning WayBlazer Travel conierge Treasure Island Customer service Ross Legal research Up All Night Music concierge 32
Thank You 33