演講 Slow Intelligence Systems 者

演講 Slow Intelligence Systems 者： - A New Approach for 張系 Component-based 國 Software Engineering 教授 1

Prof. S. K. Chang 演講者 —張系國 § 知識系統學院創辦人 (Founder, Knowledge Systems Institute www. ksi. edu)。 § 旅美教授，任教於匹茲堡大學（ Professor, University of Pittsburgh www. pitt. edu）。 § 張教授除了是電腦科學學者外，亦從事小說創作 (Writer and novelist)。 § www. cs. pitt. edu/~chang 2

Ou t l i n e v. IT does not matter! v. What is the matter? v. What is “W-H-A-T”? v. Enabling Technologies v. Slow Intelligence Systems v. SIS Applications v. Q & A 3

§ In 2003, Nicholas Carr wrote an interesting article in Harvard Business Review. Its title is: “IT Doesn’t Matter” § He argued that information technology is no longer the decisive factor in business. This article caused quite a stir. A lot of IT gurus, including Bill Gates, argued against Carr’s view. § If IT does not matter, WHAT is the matter? 4

What is IT? IT=? Information Technology 5

If IT does not matter, WHAT is the matter? 6

What is the matter? Let us return to the future. … 7

What is “W-H-A-T”? Weisure W Warfare? T AH AH Healthcare Training Agriculture Amusement? 8

W-H-A-T is in common? • • • Connected Multiple sourced Knowledge-based Personalized Hybrid 9

Smarter Planet • We are all now connected - economically, technically and socially. Our planet is becoming smarter via integration of information scattered in many different data sources: from the sensors, on the web, in our personal devices, in documents and in databases, or hidden within application programs. Often we need to get information from several of these sources to complete a task. Examples include healthcare, science, the business world and our personal lives. (Quoted from Josephine M. Cheng, IBM Fellow and Vice President of IBM Research) 10

(courtesy of IBM) 11

Hybrid Intelligence • While processor speed and storage capacity have grown remarkably, the geometric growth in user communities, online computer usage, and the availability of data is in some ways is even more remarkable. Hybrid Intelligence offers great opportunities we have to harness this data availability to build systems of immense potential. While today s large scale systems are evolutionarily based on the distributed computing technologies envisioned in the 70 s and 80 s, sheer scaling has led to many unanticipated challenges. (quoted from Alfred Z. Spector, Vice President, Research and Special Initiatives, Google, USA) 12

Hybrid Intelligence Users and computers doing more than either could individually (quoted from Alfred Z. Spector, Google). 13

Enabling Technologies Wireless Communication & Networking 14

Enabling Technologies Mobile Knowledge Agents 15

Enabling Technologies Embedded Systems 16

Enabling Technologies Distributed Multimedia Systems 17

Enabling Technologies Knowledge Based Software Engineering 18

Slow Intelligence Systems • • Slow Intelligence Systems are generalpurpose systems characterized by being able to improve performance over time. A slow intelligence system is a system that (i) solves problems by trying different solutions, (ii) is contextaware to adapt to different situations and to propagate knowledge, and (iii) may not perform well in the short run but continuously learns to improve its performance over time. 19

Slow Intelligence Systems • Slow Intelligence Systems are generalpurpose systems characterized by being able to improve performance over time through a process involving • Enumeration 20

Slow Intelligence Systems • Slow Intelligence Systems are generalpurpose systems characterized by being able to improve performance over time through a process involving • Enumeration • Propagation 21

Slow Intelligence Systems • Slow Intelligence Systems are generalpurpose systems characterized by being able to improve performance over time through a process involving • Enumeration • Propagation • Adaptation 22

Slow Intelligence Systems • Slow Intelligence Systems are generalpurpose systems characterized by being able to improve performance over time through a process involving • Enumeration • Propagation • Adaptation • Elimination 23

Slow Intelligence Systems • Slow Intelligence Systems are generalpurpose systems characterized by being able to improve performance over time through a process involving • Enumeration • Propagation • Adaptation • Elimination • Concentration 24

Slow Intelligence Systems • Slow Intelligence Systems are generalpurpose systems characterized by being able to improve performance over time through a process involving • Enumeration • Propagation • Adaptation • Elimination • Concentration • Slow Decision Cycle to complement Fast Decision Cycle 25

Slow Intelligence Systems • A SIS continuously learns, searches for new solutions and propagates and shares its experience with other peers. • From the structural point of view, a SIS is a system with multiple decision cycles such that actions of slow decision cycle(s) may override actions of quick decision cycle(s), resulting in poorer performance in the short run but better performance in the long-run. 26

SIS Basic Building Block (BBB) 27

Mathematical Formulation of BBB • For the two-decision-cycle SIS, the basic building block BBB can be formulated methematically as: if timing-control(t) == 'slow' then y(t)solution = gconcentrate (geliminate (gadapt (genumerate (x(t) problem)))) else if timing-control(t) == 'quick' then y(t) solution = fconcentrate (feliminate (fadapt (fenumerate (x(t) problem)))) 28

Advanced Building Block (ABB) 29

SIS built from BBBs and ABBs 30

OUR RESEARCH AGENDA A Framework to study Natural Slow Intelligence Systems A Test bed to develop Artificial Slow Intelligence Systems Component based Multiple decision cycles Evolutionary ontology Learning rules Visualization 31

The SIS Testbed 32

SIS Application to Product Configuration Production of personalized or custom-tailored goods or services to meet consumers' diverse and changing needs “Like its driver each Toyota Echo is unique!” 33

System Architecture 34

Ontological Transformations User Layer Instance Functionality Layer Components Layer 35

Sequence of Ontological Transformations In this way, the configuration problem CP can be formulated in its general formulation as the composition of ontological transformations: FC(FEL(FA(FEN(UR, UP)))). Similar to a SIS, the proposed Configurator can follow a slow and a fast process of solution inference. So, the previous formulation can be defined as the slow process, while the fast process can be defined as a simplified sequence of ontological transformations: FC(FEN(UR, UP)). 36

A Scenario • • • A customer would like to buy a Personal Computer in order to play videogames and surf on the internet. He knows that he needs an operating system, a web browser and an antivirus package. In particular, the user prefers a Microsoft Windows operating system. He lives in the United States and prefers to have a desktop. He also prefers low cost components. 37

Ontology for Product Configurator 38

Experimental Results • • • A set of common computer configurations based on usage scenarios were identified for evaluation The allowed configurations for the personal computer are so named: Play_Videogame (PV) Web_Surfing (WS) Online_Gaming (OG) Multimedia_Design (MD) Computer_Aided_Design (CAD) Music (MUS) Word_Processing (WP) School_Work - Web_Surfing and Word_Processing (SW) 39

User Satisfaction Index after 50 similar requests OEI (obtained FE EEI OSI I after 50 requests) USI (avg. value obtained after 50 requests) Store Reques t Italian PV 1. 0. 9 00 4 0. 8 5 0. 14 0. 87 America n WP 1. 0. 8 00 6 0. 8 3 0. 11 0. 88 British CAD 0. 0. 8 96 8 0. 7 9 0. 17 0. 83 1. 0. 9 00 0 0. 9 3 0. 07 1. 0. 9 00 2 0. 9 1 0. 06 1, 2 MD 0, 8 OSI Indian 1 0, 6 0, 4 Japanes e WS 0, 2 0 1 3 5 7 9 11 13 15 17 19 OSI Evolution Index 21 23 25 Id_Request 27 29 Italian American 0. 97 British Indian 31 33 35 37 39 41 43 45 Japanese 47 0. 98 49 40

SIS Application to Detect Trends/Topics 41

SIS module 1 42

SIS module 2 43

Discussion • There a large number of intelligent systems, quasi-intelligent systems and semi-intelligent systems that are "slow". Distributed intelligence systems, multiple agents systems and emergency management systems are mostly slow intelligence systems that exhibit the characteristics of multiple decision cycles. 44

Discussion (continued) • Since time is relative, "slow" intelligence systems for some can also be "fast" for others. • A slow intelligence system can evolve into a fast intelligence system. • A SIS differs from expert systems in that the learning is not always obvious. 45

Conclusions • In the age of micro-profit economy, Information Technology to acquire, communicate and apply knowledge to reduce cost and improve efficiency will still be a decisive factor. • IT is KNOWLEDGE TECHNOLOGY. • EVERY INDUSTRY is IT INDUSTRY. 46

What is the matter? The future is already here! 47

Q&A 48

The End 49

演 講 Slow Intelligence Systems 者 —

演講 Slow Intelligence Systems 者 —