SANJEEV GUPTA Addressing the issues of KM for

Скачать презентацию SANJEEV GUPTA Addressing the issues of KM for

704e7df492eb43ee2edf841903027ac3.ppt

Количество слайдов: 54

IBM Research – India What is Accessibility? Access to information technology regardless of ability or disability “Accessibility – which started out as a philanthropic effort – has now evolved to a business transformation effort for IBM and our clients. ” Sam Palmisano, IBM CEO, April 27, 2004 3 © 2009 IBM Corporation

IBM Research – India IBM Vision on Accessibility From compliance to societal transformation. . . IBM’s vision is to enhance human capabilities through technological innovation so that societal participation and personal fulfillment can be maximized, regardless of age or ability…. Accessibility is not about “them”, it’s about ALL of us… 4 © 2009 IBM Corporation

IBM Research – India Accessibility in India – Broader Landscape § Factors affecting accessibility – Physical disability • About 60 million people with disability • 42. 5% of the disabled population comprises women • 75% of persons with disabilities live in rural areas – Educational/Economic disability • About 70% of India’s population lives in villages • A large percentage of them are poor • Literacy is mainly limited to write their names, make signatures, read large hoardings 5 – Low internet penetration • Computer penetration is still very low • People in remote areas not very comfortable in using computers © 2009 IBM Corporation

Addressing India Accessibility Issues IBM Research – India Ø EWB ØHindi Speech Ø Web. Adapt 2 M Recognition ØIndian English TTS e Ø a. Designer Ma Ac king ce W ss eb ibl e IBM Research Innovation e th e g d gin divi id Br ital dig Novel ways for Information sharing ØTelecom Web/ WAV 6 © 2009 IBM Corporation

The Spoken Web IBM Research - India 8

Introducing Voice. Sites « A Voice. Site is: ª ª Consists of a set of interconnected Voice. Pages (eg vxml files) ª Accessed by calling up the associated phone number and interacting with its underlying application flow through a telephony interface ª IBM A voice driven application hosted in the network and created by subscribers themselves Analogous to websites in the World Wide Web

Voi. Gen Call Voi. Serv to create Voice. Site Voi. Serv You have now created your voice site successfully. to publish Pleaseenterthe name can Do you like say accept profession Please say yourjobsprovide Wouldspecify name and now Wouldyour to. Users Please you like to your Please record likewelcome Voi. Serv: Wouldspecify your offerwhile access number appointment voiceinforyour phone hours you service charges of site through your information your some references yellow home location workingare away from messageyour schedulingyour phone services? number. Thank you for using references home pages? location? work? this system. Caller: Voice. Site Calendaring Service Database Yellow Page Service Location Tracker IMS Presence Server IBM You to 5 dollars an about Yes can 7 pmto Jack hour Hi my Delhi is Sam, and I am a South I am 9 charge Plumber talk Sam name my work. His number plumber. Please find is 41292100 information regarding my services on my Voice. Site. Yellow Pages Server WWW Yellow. Pages Website

Small user study – Plumber Voice. Site Carpenters/ Electricians make a call to Voi. Gen to generate their voice sites « Methodology for survey ª ª « Electricians/Plumbers/Carpenters make a call to Voi. Gen and create their voice sites We ask the subjects about the usability of the Voi. Gen system 12 subjects surveyed for technology validation ª 10 were able to create the voice site successfully (within 4 minutes) ª There were usability issues with respect to conversation flow, speech recognition accuracy Everyone realised that this technology can have tremendous impact Since this technology does not require the end-user to own any costs in terms of devices, it has a low acceptance barrier ª ª IBM The voice sites are automatically deployed in the system People call these voice sites to schedule time with the specialists

What is the Spoken Web? The Spoken Web is a world wide web in the telecom network, where people can host and browse Voice. Sites, traverse Voi. Links, even conduct business transactions, all just by talking over the existing telephone network. ª The Spoken Web will interoperate with the existing WWW. Spoken Web will interoperate with Next Generation Networks too. IBM

Spoken Web enables multiple business opportunities « New source of revenue opportunity for telecom operators Creation and hosting voicesites ª Payments and financial transactions ª « SMBs and microbusinesses can leverage the T-Web « Examples Microbusiness Voicesite ª Voice. Site Personalisation ª Rural Voikiosk ª “Anyone with a mobile handset can become a T-Web enabled microbusiness voicesite owner and accessor, and also conduct transactions on the T-Web” IBM

Spoken. Web Andhra Pilot Statistics « « Pilot Launch: May 23, 2008 Report Summary (ended on Jan 28, 2009) ª ª ª ª ª IBM Total number of calls received = 114782 Number of unique callers = 6509 Total time spent = 2135 hours Average call time spent = 0 hours, 1 min, and 14 seconds. Maximum call duration = 0 hours, 49 min, and 40 seconds. Minimum call duration = 0 hours, 0 min, and 0 seconds. Number of calls to Ashwini Center = 8399 Number of calls to Health Center = 14216 Number of calls to V-Agri = 13881 Number of calls to Professional Services = 37112

Motivation § Web is a rich source of useful local information Ø Weather, travel, entertainment, insurance, finance § However a significant population (specially in emerging countries) is not using this information due to Ø computer skills, exposure to browsing, language skills, physical limitations, aging § A large number of such people have access to phone (landline/mobile) Ø Growing at a fast rate § Even computer users can’t browse in several conditions Ø 16 On the move, no connectivity areas, low speed, etc. © 2006 IBM Corporation

Proposition § Decouple web information from web browsing Ø Let the people access web information without having to browse/know how to browse Ø However, still leverage the web interface • No change required on the website/content provider side § Let the system browse instead of the user Ø System figures out how to extract the information from web for a user’s query § The interaction can be enabled in user’s language for simple queries (structured input/output) Ø 17 through speech recognition and language translation © 2006 IBM Corporation

Scenarios § A person wants to go from station A to B. He wants to know what all trains are available, their schedule, availability, etc. He has only a phone and is not familiar with the web. Ø Access a relevant website (e. g. , indianrail. gov. in) • Get the required inputs from the user – Source, destination, class, dates, etc. • Fetch the information from the web and give it back to the user § A person wants to know what are the interest rates offered by various banks for home loan Ø System can goto a popular website (e. g. , apnaloan. com) • Get the required inputs from the user – Term, floating, fixed • Fetch the information from the web and give it back to the user – lowest interest rate offered § A person is planning a trip to Chennai and wants to know the current weathere Ø Goto cnn. com • Fill up the form • Speak the weather over phone in local language Ø Goto google • Get the weather information and reply back to the user 18 © 2006 IBM Corporation

What is currently available? § Web browsers on mobile phones Ø Person can browse the web on handheld device • Costly, complicated, tiny interface, inaccessible, not suited for common man, § Browsing of voice sites Ø Created from scratch using VXML § Speech interface to specific services (Tell. Me, Nuance) Ø Nearestaurant, police station, hospitals, etc. • • • Based on knowledgebase created offline OR Proprietary tie-up with content providers to have access to databases Predefined, Keyword based A third-party data provider gathers the business information that Tellme provides in the Tellme download and on 1 -800 -555 -TELL, so we are unable to directly add or correct specific business information. If you would like to add or correct information that is listedfor your business, please use the easy form on the Info. USA website. (Taken from Tell. Me Website : http: //www. tellme. com/you/faqs) 19 © 2006 IBM Corporation

How does it work? Service 1 Service 2 Web Site Service 3 Service 4 Response Process Generator Information Extraction tools Request Process Generator Browser scripting tools such as Co-scripter 21 © 2006 IBM Corporation

Request Generation & Execution § Leveraging the browser interface through scripts Ø Ø Inputs 22 Generate a script with inputs taken from the user Execute the script with a browser Input Collection (VXML) Data for Scripting Tool Data Web Browser Web-Page Script © 2006 IBM Corporation

Information Extraction & Response § Use HTML Syntax and Semantics to extract information Ø Look in HTML sections using syntax knowledge Ø Use semantics based on context and keywords Information Extraction Module Web-Page’s HTML Source Relevant Keywords/ Semantics Syntax Request Generation & Execution Response User Interaction (Iterative) Keywords : Airline / Lowest / Cheapest / Prices HTML Syntax : TABLE , ROW-COLUMN () 23 User Top three cheapest flights are : Go Air 4435 Rs at 5: 05 AM Deccan 4449 Rs at 4: 15 AM Spice Jet 4859 Rs at 8: 00 AM © 2006 IBM Corporation

Overview • Having difficulty viewing Web pages? Easy Web Browsing is a solution that helps bridge the digital divide for novice computer users, people who are experiencing vision loss, second-language learners, seniors, and persons with reading challenges • Highlights – – – Installs by automatically downloading from Web site. Reads text aloud with adjustable speed and volume control. Allows users to customize size and color of Web content. Ruler function that helps users find and follow their reading position. Highlight function focusing on the reading text with four patterns of marking. – Customizable line and word spacing features that enhance readability. 25 © 2008 IBM Corporation

Summary • • Financial, Retailers, Travel and Government Industries are interested in Web & Kiosk accessibility. • IBM’s EWB is a quick and reasonable solution to make web sites and Kiosk more accessible for consumers, citizens and travelers. • It also drag new opportunities by combined offering • 27 Accessibility web sites are required for PWD’s but they offer seniors, novices and non native speakers assistance as well. Both the customers and the end users are satisfied with this solution © 2008 IBM Corporation

Reading Companion “Reading Companion has opened new cultural horizons for our children. With such a wide choice of books to increase their vocabulary and improve their comprehension skills, they’re developing a true love for reading. ” readingcompanion. org Patricia Diaz Covarrubias, Executive Director, Christel House de Mexico, A. C. IBM’s multi-million-dollar investment in literacy, using voice recognition technology over the web to help children and adults learn to read. – Anytime, anywhere web access, providing feedback and asneeded assistance – More than 1, 380 schools and nonprofit organizations -- about half of which are schools -- in 25 countries and approximately 56, 200 users are participating in this grant program. 850 schools & nonprofit organizations in 26 countries, benefitting more than 40, 000 children and adults – Evaluation showed: • Child: higher test scores on word recognition and reading comprehension • Adult: Increased English communication skills and literacy; positive job outcomes for some learners 28 © 2010 IBM Corporation

IBM Research IBM: Employing Diversity & Excellence Meet IBMer Dimitri Kanevsky: • • Master Inventor in IBM Research • 2002 Science Accomplishment for Maximization Algorithms • 29 Deaf Generated 80 IBM patents © 2009 IBM Corporation

IBM Research IBM: Employing Diversity & Excellence Meet IBMer Mike Squillace: • • Joined IBM in 2002 • Sun Certified Java Programmer • Ph. D in Philosophy and B. S in Computer Science • 30 Blind Developed Patents for multiple GUI architectures and defining GUIs via mark up languages & reflection © 2009 IBM Corporation

IBM Research IBM: Employing Diversity & Excellence Meet IBMer Chieko Asakawa: • • Joined IBM Research in 1985 • An IBM Fellow • Member of Women in Technology Hall of Fame • 31 Blind Developed Digital Braille System & 3 key applications © 2009 IBM Corporation

IBM Research Recognition by the Hon President of India in 2007 & 2009 for providing technology for people on the other side of the digital world to make complete knowledge society. 32 © 2009 IBM Corporation

IBM Research a. Designer • Characteristics - Visualization of blind usability - Simulation of low-vision users’ view - Weak eyesight, color vision deficiency, cataracts. - Checking compliance items - WCAG, Section 508, IBM CI 162, JIS, etc. • Award - Wall Street Journal Technology Innovation Award 2004 (Runner-up) • Status - Opensourced as a basis of Eclipse. org ACTF (Accessibility Tools Framework) 34 © 2009 IBM Corporation

IBM Research Blind Usability Visualization Example Original With heading Tags Inaccessible With skip-link Easy to find main contents 35 • Headers can use as TOC • Easy to navigate through the page © 2009 IBM Corporation

IBM Research Low Vision Simulating the experience of users who have low vision Low vision simulation. In this example, Color Vision Deficiency (Deutan) and cataract are simulated. The original Web page which people without low vision view. Problem map that indicates the positions of problems. Summary Report Setting panel (Eyesight, color vision deficiencies, crystalline lens transparency) 36 © 2009 IBM Corporation

IBM Research Overview Famous for its science and vital to IBM 1952 San Jose California Almaden 1986 1945 1 st IBM Research Lab in NY (Columbia U) Watson 1961 Innovation that Matters Zürich 1955 Beijing 1995 Business New Insights Austin 1995 1970's Haifa 1972 Tokyo 1982 1980's ♦ Corporate funded research agenda ♦ Collaborative team ♦ Technology transfer ♦ Effectiveness Centrally funded 37 ♦ Shared agenda Joint programs Society Delhi 1998 1990's ♦ Work on customer problems 2000's ♦ Create business advantage for customers e. Business research Research in the marketplace Technology ODIS On Demand Innovation Services EBO Emerging Business Opportunities FOAK First of a Kind Technology Transfer © 2009 IBM Corporation

IBM Research Focus Areas Business Areas Service Delivery Infrastructure Services Application Services Emerging Solutions Contact Center Services Telecom Software Others (Banking, etc. ) Systems Technical Competencies Computer Science Service Science • • • 40 Math Science • • • Distributed Systems – system mgmt. , middleware Information Management – data mining, machine learning Interaction Technologies – speech Programming Technologies – parallel and hi-perf. prog. Software Engineering – model-driven, distributed dev. Operations Research Algorithms Optimization Game Theory Service Engineering Service Productivity Service Management Service Quality Service Supply Chains © 2009 IBM Corporation

IBM Research Easy Web Browsing – UI technologies (1) § Easy operations and Easy-to-use operation panel – No URL input field, could only surf within specified domains. – Operation Panel • Navigation (Home/Back/Stop) • Voice speed/volume • Zoom • Line Spacing • Ruler • Color setting • Print • Detail Setting • Help 42 © 2009 IBM Corporation

IBM Research Easy Web Browsing – UI technologies (2) § Read aloud with speed control § Character enlarging (/w screen magnifier) Accessibility at IBM means enabling IT hardawa, 43 © 2009 IBM Corporation

IBM Research Easy Web Browsing – UI technologies (3) § Background color change –Color vision deficiency –Cataract –Weak sighted Black text on a white background with blue for links for normal display 2. Yellow text on a blue background with white for links 3. Black text on a light yellow background with blue for links 4. 44 1. Yellow text on a black background with white for links © 2009 IBM Corporation

IBM Research Easy Web Browsing – UI technologies (4) § Automatic language switch (panel, TTS etc) according to the lang attribute of the Web page. – Support for thirteen languages : Chinese (Simplified), Chinese Traditional (Taiwan, Hong Kong), English (US and UK), French, German, Italian, Japanese, Korean, Spanish, and Portuguese (Brazil, Portugal). Japanese English 45 © 2009 IBM Corporation

IBM Research Sensei § An automated tool for assessing spoken English skills – Evaluates pronunciation, grammar, comprehension – Uses advanced speech processing techniques – Provides scores for each of the categories in real time § The tool is Web enabled – Can be used for remote hiring/assessment – Can be used for training – Centralized database/content update § Can help children learn English language 47 © 2009 IBM Corporation

IBM Research Evaluation of Syllable Stress § Lexical stress evaluation – Important for spoken English comprehension • Meaning changes with stress pattern (PROject, pro. JECT, con. TENT, CONtent) – Different stress point for different words (a. VAilable, Industry) § Primary features – pitch, duration & energy § Challenges – Every word has a different stress pattern – Stress can also change depending upon context – Relative importance of the features varies for different words and speakers – primary syllable can be inherently low in energy or short in duration § Word Dependent Classifiers – A separate classifiers is trained for each of the words – Performs better than the word independent models § Single classifiers – Estimating the multi-dimensional shape corresponding to the correct class (spanned only by the correct utterances) § Word Independent Classifiers – Classify each individual syllables into stressed/unstressed – Combine soft decisions to determine correctness at word level 48 Acoustic Feature Used (Syllable level) General behavior (if stressed) 1. Average Fundamental Frequency (F 0) Higher 2. Average energy higher 3. Duration Longer 4. Average filtered energy (above 4 k. Hz) Higher 5. Average energy * duration Higher 6. F 0 * duration higher 7. F 0 ratio (of next syllable to current) Lower 8. Accuracy with Human Energy ratio lower (of next syllable to • Human Assessors Repeatability is 86% current) • Human Assessors Reproducibility is 64% • Human Assessors Accuracy is 85% • Sensei Accuracy is 81 % • Sensei Reproducibility is 100% © 2009 IBM Corporation

IBM Research Evaluation of Spoken Grammar Possible responses • • Prompt: Both the dogs is barking Correct answers: § Both the dogs are barking § Both the dogs were barking Grammatically correct or incorrect sentence Candidate records correct sentence Both the dogs is barking (x) Both the dogs are barking (b) Those dogs are barking () Both the dogs were barking (b) Speech Recognition 1: assigned for correct sentence § Evaluate spoken grammar skills of the candidate – Not possible to evaluate free speech – low recognition accuracy, LM bias – Prompts and answer (make it interactive) – Prompts designed to test various parameters • Tenses, articles, propositions, subject-verb agreement § Challenges – Correct and incorrect answers acoustically close to each other – Multiple correct answers are possible – Incomplete recordings (last word chopped), response outside speech grammar – Content challenges 49 • Effectiveness of questions 0: assigned for incorrect sentence Accuracy with Human • Human Assessors Repeatability is 94% • Human Assessors Reproducibility is 82% • Human Assessors Accuracy is 95% • Sensei Accuracy is 85 % • Sensei Reproducibility is 100% © 2009 IBM Corporation

IBM Research Evaluation of Articulation Decision Level Accuracy with Human § Impact Sounds – S, sh, z, sh, v, w, t, d, AO (ball) – Correct pronunciation of words § Different from speech recognition – Recognition should discard pronunciation variation § Customization of acoustic models – Models trained from models speakers – US/UK models adapted to model speakers – Indian English models adapted to model speakers § Features – phone confidence scores. – Word endings – Duration of phones Accuracy with Human § Challenges – Subjectivity in human ratings – Lack of model speakers data (only few model speakers) – Other considerations in human rating – stress, fluency, etc. 50 © 2009 IBM Corporation

IBM Research Combined Scores by Sensei and Assessors Correlation between Sensei and Average Assessor Score = 0. 80 Correlation between individual assessors Assessor 1 Assessor 2 Assessor 3 Sensei 0. 78 0. 75 0. 79 Assessor 1 - 0. 89 0. 93 - 0. 90 Assessor 2 Assessor 3 51 - © 2009 IBM Corporation

IBM Research Social Accessibility New methodology to make webpages on the Internet more accessible by gathering users’ voices and by using the power of the open community. Social computing + accessibility Any Web user can improve accessibility of any webpage on the Internet without changing the original content. Any visually impaired user can join the improvement process through various collaboration mechanisms. 1) Encounter a problem in Web content. Report improved page! 54 Notification Social Accessibility Service Visually impaired users “Ye s we can ” 3) Access the 2) Respond by using authoring tools. Accessibility metadata Load Submit Sighted volunteers Generate Accessibility metadata alttext = “Yes we can” © 2009 IBM Corporation