Seminar
9 likes•4,825 views
1) Intelligent software assistants are agents that can perform tasks with minimal direction from users by interacting through natural conversation. 2) Siri was an early virtual personal assistant that could answer questions, make recommendations, and perform actions through natural language. 3) Developing intelligent assistants that can interact with the world, users' friends and contacts, and the user themselves remains a challenge and area of ongoing research.
Seminar
- 1. Intelligent Software Assistants Meghaditya Roy Chaudhury Exam Roll ~ CSE-1241040 Class Roll ~ 000810501052 BCSE Final Year Seminar-I Jadavpur University
- 2. The Future of computing and AI “ Our ultimate objective is to make programs that learn from their experience as effectively as humans do. We shall…say that a program has common sense if it automatically deduces for itself a sufficient wide class of immediate consequences of anything it is told and what it already knows.” -John McCarthy, "Programs with Common Sense", 1958.
- 3. What are Intelligent Software Assistants? It is a software agent which performs a task with minimum specific directions from users. What we do now – - Follow links on search results. What we will do – - Interact by having a conversation. MIT Technology Review(2009 TR10 special reports) “ one of the 10 emerging technologies of the decade” ( http://www.technologyreview.com/specialreports/specialreport.aspx?id =37 )
- 4. Siri : Virtual Private Assistant The application uses a natural language user interface to - answer questions. - make recommendations - perform actions
- 5. Why was it necessary? Siri was founded in 2007 by - Dag Kittlaus - Adam Cheyer - Tom Gruber - Norman Winarsky Adam Cheyer: "Here's what I want to do, you programs figure out who needs to be involved."
- 6. 3 main technical components A conversational interface Personal context awareness Service delegation
- 7. VPAs perform threefold tasks Interacting with the World. Interacting with the User’s friends and contacts. Interacting with the User.
- 8. Hard Part: Interacting with the World
- 9. Search is not enough! The principle goal of Web Intelligence - Add Deduction Capability to Search Engines. Search Engine = (Identification of Query-relevant information + relevance-ranking of query-relevant information). QA system = (Search Engine + deduction from Query-relevant information)
- 10. Test Query (Google) What is the distance between the largest city in Portugal and the largest city in Spain? ---- Failure! What is the largest city in Portugal? -- Lisbon (Success) What is the largest city in Spain? -- Madrid (Success) What is the distance between Madrid and Spain? -- 502.84 km (Success)
- 11. From a search-engine to a do-engine The user can ask a broad question It pulls information from the web It interprets this reply The software offers specific suggestions By now, the conversation has narrowed to do is click on his choice.
- 12. Tasks performed by a QA System Question. Question Classification. Information Retrieval. Answer Extraction. Answer
- 14. Harder Part: Interacting with the user’s friends and other contacts
- 15. Emerging community-based QA system such as “Quora”
- 16. Example of Social Search : iLink This is a general interaction model - social search and message routing. FAQ generation in a social network (FAQtory). The underlying theory involves - Markov Decision Processes - The Pagerank model.
- 17. Hardest Part: Interacting with the User
- 18. Semantic Desktop Framework: I ntegrate .R elate .I nfer .S hare . It enables users to create a “personal map” across their office-related information objects. It provides - dashboard views. - contextual navigation. - relationship-based structure across office applications.
- 19. Integration of Applications Information resources (e.g., an email message, a calendar appointment). A knowledge base (KB) provides the unified data model, persistence store, and query mechanisms. The IRIS user interface framework.
- 20. Open Agent Architecture Interactions among software components Flexible Co-operative. Some of the most important UI qualities include: Dynamic : Adaptable display : Multimodal : Multi-user Collaborative : Social :
- 21. Virtual Assistant: The time is now! Increasing amount of structured data on the Web. Wide availability of open API’s for web services. Global adoption of the Web. Growing adoption of 3G smart-phones and tablets.
- 22. Interesting Application Areas Medical Diagnosis Assistant. Assistance for visually impaired persons. IVRS for customers. Ex-Panasonic’s Automated System for helping customers. Speech Recognition Systems enabled with GPS for cars. Ex-Ford’s Sync.
- 23. Research Work : CALO (Cognitive Agent that Learns and Organizes) Participating Organizations and Universities 1. Carnegie Mellon University 2. Georgia Institute of Technology 3. Harvard University 4. Massachusetts Institute of Technology 5. Stanford University 6. State University of New York - Stony Brook 7. University of California at Berkeley 8. University of Massachusetts at Amherst 9. University of Michigan 10. University of Pennsylvania 11. The University of Texas at Austin 12. University of Washington 13. Yale University
- 24. Last Note! “ If you invent a breakthrough in artificial intelligence, so machines can learn, that is worth 10 Microsofts” . – Bill Gates, quoted by NY Times (http://www.nytimes.com/2004/03/01/technology/01bill.html)
- 25. Thank You