Smart B2C eBusiness The Employment of Intelligent Agent Technology

1
Smart B2C e-Business- The Employment of
Intelligent Agent Technology

• Alan Aizhong Lin Mao Lin Huang

2
The evolution of e-Business --- Online e-Business
Online e-Business (Software components support
human beings to achieve business
goals) Support human beings to advertise
products Support interactions between human
beings Manage processes for human beings
3
The evolution of e-Business --- Smart e-Business
Smart e-Business (Software components
achieve business goals autonomously) Advertise
products Interact with other components Harvest
process knowledge from business processed
4
Outline

• Motivations A Grocery Shopping-Cart Agent Demo
• Software Agent
• Definitions
• Building Software Agent
• Architecture reactive, proactive (BDI), Hybrid,
  Hybrid and Layered
• Multi-Agent System (MAS)
• Architecture
• Agent Communication Protocols
• Agent Interactions
• Self-interested Agent interactions
• Auction
• Auction-English
• Auction-Dutch
• Auction-Vickrey
• Negotiation
• Contract-Net
• References

5
Motivation --- A Grocery Shopping Agency

• Human beings, sometimes, are busy or lazy. They
  are not going to shop goods or foods in person
  (including online shopping). They employ software
  agents to find and buy the cheap, healthy and
  appropriate goods or foods.
• For example, we want to have a lunch, but we
  dont want to spend time in choosing and
  purchasing goods. Instead we assign this task to
  a software agent.
• A grocery shopping agency demo from
  AgentBuilder http//www.agentbuilder.com/AgentTe
  chnology/agencyviewer.html

6
The Grocery Shopping Agency --- 1
7
The Grocery Shopping Agency --- 2
8
The Grocery Shopping Agency --- 3
9
The Grocery Shopping Agency --- 4
10
The Grocery Shopping Agency --- 5
11
The Grocery Shopping Agency --- 6
12
The Grocery Shopping Agency --- 7
13
The Grocery Shopping Agency --- 8
14
Software Agent --- definition --- 1

• An agent is physical or virtual entity that can
  be viewed as perceiving its environment through
  sensors and acting upon that environment through
  effectors.

• An intelligent agent (or rational agent) is an
  agent that takes rational actions to meet the
  design objectives.
• Artificial Intelligence A Modern Approach by
  Stuart Russell and Peter Norvig, c 1995
  Prentice-Hall, Inc.

• Wooldridge defines an intelligent agent is an
  encapsulated computer system, situated in some
  environment, and capable of flexible autonomous
  action in that environment in order to meet it
  design objectives Wooldridge99.

15
Software Agent --- Definition --- 2

• Artificial Intelligence A Modern Approach by
  Stuart Russell and Peter Norvig, c 1995
  Prentice-Hall, Inc.

16
Software Agent --- definition --- 3

• A software intelligent agent works as a
  human-like software problem solving entity that
  is designed with the capabilities to
• live it runs to look after an environment
• see it perceives the changes (events) of the
  environment
• think (reason) it chooses actions by using its
  reasoning, decision-making and interaction
  mechanisms
• talk it interacts with other agents
• do it executes actions to respond the
  changes of the environment
• learn It learns new abilities or knowledge
  from its user or other agents

17
Software Agent --- definition --- 4

• Assume the environment may be in any of a finite
  set E of discrete, instantaneous states
• E e0, e1,
• Agents are assumed to have a repertoire of
  possible actions available to them, which
  transform the state of the environment.
• Ac a0, a1,
• A run, r, of an agent in an environment is a
  sequence of interleaved environment states and
  actions

18
Software Agent --- definition --- 5

• Let
• R be the set of all such possible runs
• RE be the subset of these that end with an
  environment state
• Then
• An Agent is a function which maps runs to actions
• Ag RE -gt Ac
• An agent makes a decision about what action to
  perform next based on the evolution of the
  environment observed to date.
• And
• An agent-based system is a pair of an agent and
  an environment
• Any agent-based system can be associated with a
  set of possible runs we denote the set of runs
  of an agent Ag in environment Env by R(Ag, Env)

19
Software Agent --- definition --- 7

• An intelligent agent could have following
  properties, but it is not required to have all of
  them.
• Autonomy Agents operate without the direct
  intervention of human or others, and have
  some kind of control over their actions and
  internal state.
• Reactivity Agents are able to perceive their
  environment and respond to it in timely
  fashion to changes that occur in it.
• Pro-activeness Agents are able to exhibit
  goal-directed behavior by taking the
  initiative.
• Social Ability Agents are able to interact with
  other agents (and possible humans) via some
  kind of agent- communication language
• Rationality Agents act in order to achieve its
  goals and do not act intentionally in such a
  way as to prevent its goals being achieved.
• Adaptability Agents can adapt themselves to new
  tasks or new environment.
• Mobility Agents can move around a network.

20
Building Software Agent

• Two key problems
• How do we build agents that are capable of
  independent, autonomous action in order to
  successfully carry out the tasks that we delegate
  to them?
• How do we build agents that are capable of
  interacting (cooperating, coordinating,
  negotiating) with other agents in order to
  successfully carry out the tasks that we delegate
  to them, particularly when the other agents
  cannot be assumed to share the same
  interests/goals?

21

• Key 1
• How an agent realizes its autonomous property?

22
Agent Architecture --- the kernel of an agent

• An intelligent agent architecture is a
  specification that describes how an intelligent
  agent derives rational actions to respond the
  events perceived from the environment to meet the
  goal
• Reactive Reasoning
• Proactive Reasoning

23
Agent Architecture --- reactive reasoning --- 1

• The reactive reasoning agent architecture pursues
  a direct way to derive the next action in any
  given state.
• Perceive environment
• Condition-Action (or Event-Condition-Action)
  rules
• Action to change environment

• An Example
• if path_clear(t) then move forward rule 1
• if !path_clear(t) then turn right rule 2

24
Agent Architecture --- reactive reasoning --- 2

• Woorldridge99
• function action(p, P) A
• begin
• fired (c, a) (c, a) in R and p in c
• for each(c, a) fired do
• if ((c, a) fired such that (c, a) lt (c,
  a))
• then
• return a
• end-if
• end-for
• return null
• end function action
• P a set of percepts, A a set of actions
• R a set pf rules
• (c, a) a rule, r1 lt r2 r1 is better than r2

25
Agent Architecture --- proactive reasoning

• Proactive reasoning provides an indirect way to
  derive the next action in any given state.
• A number of proactive approaches have emerged as
  candidates for the study of agents BIP88
  Doyle92 RG91 RK86 Shoham93.
• One such architecture views an agent as having
  certain mental attitudes of Belief, Desire and
  Intention (BDI) Bratman87RG91.
• The basic idea of the BDI approach is to describe
  the internal processing state of an agent by
  means of a set of mental categories, and to
  define a control architecture by which the agent
  selects its sequence of actions based on their
  representation.
• The mental categories are belief, desire and
  intentions which represent the information,
  motivational and deliberative states of the agent
  respectively.

26
Agent Architecture --- proactive reasoning --- BDI

• Beliefs Information about the environment
• Desires Objectives to be accomplished, possibly
  with each objectives associated
  priority/payoff
• Goals something the agent is working on or
  towards
• Intentions The currently chosen course of
  action
• Plans Means of achieving certain future world
  states. Intuitively, plans are an abstract
  specification of both the means for achieving
  certain desires and the options available to the
  agent.
• Events significant occurrences.
• From Winikoff01

27
Agent Architecture --- proactive reasoning --- BDI

• BDI-interpreter RaoGeorgeff91
• initialize-state()
• repeat
• options options-generator(event-queue)
• selected-options deliberate(options)
• update-intentions(selected-options)
• execute()
• get-new-external-events()
• drop-successful-attitudes()
• drop-impossible-attitudes()
• end repeat

28
Agent Architecture --- Hybrid

• A hybrid agent architecture combines reactive and
  proactive agent architecture

• From Winikoff01

29
Agent Architecture --- Hybrid and Layered

• A layered agent architecture provides world
  (reactive), mental (BDI), and social reasoning.
• BR P x B -gt B
• SG B x G -gt G
• PS B x G x I -gt I

• From Müller96

30
Agent autonomous property is realized

• After the agent architecture is built, when an
  agent equipped with the architecture, beliefs
  (facts and rules), and actions, it can realize
  the autonomously property

31

• Key 2
• How an agent realizes its social property?

32
Multi-Agent System (MAS)
33
MAS --- Architecture
34
MAS --- Features

• Features
• Which interact through communication
• Are able to act in an environment
• Have different spheres of influence
  (cooperative or selfish)
• Will be linked by other (organizational)
  relationships

35
MAS --- Agent Communication --- protocols
36
MAS --- Communication Protocols

• Network Layer realizing low level communication
• TCP/IP, Socket,
• Content Layer representing knowledge
• Text, KIF, Prolog,
• Message Layer wrapping the content
• KQML, FIPA ACL,
• Interaction Layer realizing high level
  conversation
• Vote, Auction, Contract Net, negotiation,
  coordination,

37
MAS --- ML --- Speech Acts

• In general, a speech act can be seen to have two
  components
• a performative verb (e.g., request, inform, )
• propositional content (e,g,, the door is
  closed)
• Examples
• performative request
• content the door is closed
• speech act please close the door
• performative inform
• content the door is closed
• speech act the door is closed!
• performative inquire
• content the door is closed
• speech act is the door closed?

38
MAS --- ML --- FIPA ACL

• The Foundation for Intelligent Physical Agents
  (FIPA) started work on a program of agent
  standards --- the centre piece is an ACL

39
MAS --- ML --- FIPA
40
Interaction Layer (IL)

• Benevolent Agents Interaction
• Task sharing and Result sharing
• Self-interested Agents Interaction
• Auction
• Auction-English
• Auction-Dutch
• Auction-Vickrey
• Negotiation
• Contract-Net

41
MAS --- IL --- BAI --- Task sharing and result
sharing

• Two main modes of cooperative problem solving
• Task sharing components of a task are
  distributed to component agents
• Result sharing information (partial results etc)
  is distributed

42
MAS --- IL --- BAI --- sharing via blackboard
systems

• The first schema for cooperative problem solving
  the blackboard system
• Results shared via shared data structure
• Multiple agents can read and write to blackboard
• Agents write partial solution to blackboard
• Blackboard may be structured into hierarchy

43
MAS --- IL --- Auction

• An auction takes place between an agent known as
  the auctioneer and a collection of agents known
  as the bidders
• The goal of the auction is for the auctioneer to
  allocate the good to one of the bidders
• In most situations, the auctioneer desires to
  maximize the price bidders desire to minimize
  price.

44
MAS --- IL --- Auction Parameters

• Goods can have
• private value, public/common value, correlated
  value
• Winner determination may be
• first price, second price
• Bids may be
• open cry, sealed bid
• Bidding may be
• one shot, ascending, descending

45
MAS --- IL --- Auction-English

• The most commonly known type of auctions with
• first-price
• open cry
• ascending
• Dominant strategy is for agent to successively
  bid a small amount more than the current highest
  bid until it reaches their valuation, then
  withdraw
• The good is allocate to the agent who bids the
  highest price
• An agent decides the bid price by using

46
MAS --- IL --- Auction-English protocol
47
MAS --- IL --- Auction-Dutch

• A type of auctions with
• open cry
• descending
• Auctioneer starts an opening price at an
  artificially high value
• Auctioneer lowers offer price until some agent
  makes the bid equal to the current offer price
• The good is allocate to the agent that made the
  offer
• An agent decides the bid price by using

48
MAS --- IL --- Auction-Dutch
49
MAS --- IL --- First-price Sealed-bid auction

• A type of auctions with
• One-shot
• First price
• Sealed-bid
• Features
• There is a single round
• Bidders submit a sealed bid to the good
• Good is allocated to agent that made highest bid
• Winner pays price of highest bid
• Best strategy is to bid less than true valuation
• An agent decides the bid price by using b (?µ
  )/2

50
MAS --- IL --- Vickrey auctions

• A type of auctions with
• Second-price
• Sealed-bid
• Good is awarded to the agent that made the
  highest bid, at the price of the second highest
  bid
• Bidding to your true valuation is dominant
  strategy in Vickrey auctions
• An agent decides the bid price by using ß O (x)
  x.

51
MAS --- IL --- Negotiation

• Auctions are only concerned with the allocation
  of goods richer techniques for reaching
  agreements are required.
• Negotiation is the process of reaching agreements
  on matters of common interest.
• Any negotiation setting will have four
  components
• A negotiation set possible proposals that agents
  can make
• A protocol
• Strategies, one for each agent, which are private
• A rule that determines when a deal has been
  struck and what the agreement deal is.
• Negotiation usually proceeds in a series of
  rounds, with every agent making a proposal at
  every round.

52
MAS --- IL --- A scenario of Negotiation

• Your company have a building construction
  contract, but your company can not complete it
• Your company announces sub contracts to other
  companies
• Your company and other companies can discuss and
  come to an agreement that is better for both of
  you

53
MAS --- IL --- Negotiation --- The Contract Net

• Well known task-sharing protocol for task
  allocation is contract net
• Recognition
• Announcement
• Bidding
• Awarding
• Expedition

54
MAS --- IL --- Negotiation --- The Contract Net
55
MAS --- IL --- Negotiation --- The Contract Net
--- protocol
56
E-Business An intelligent agent application area

• An important rationale for intelligent agents is
  the potential for electronic business
• Most commerce is currently done manually. But
  there is no reason to suppose that certain forms
  of commerce could not be safely delegated to
  agents
• A simple example finding the cheapest copy of
  Office 2000 from online stores
• More complex example flight from Sydney to Perth
  with window seat and does not use a fly-by-wire
  control
• Simple examples
• first-generation e-commerce agents
• Bargain Finder from Andersen
• Jango from NETBOT (now EXCITE)
• Second-generation negotiation, brokering,
  market systems

57
References

• http//www.agentbuilder.com/AgentTechnology/agency
  viewer.html
• Artificial Intelligence A Modern Approach by
  Stuart Russell and Peter Norvig, c 1995
  Prentice-Hall, Inc.
• Michael Wooldridge. Intelligent Agents,
  Multiagent Systems A Modern Approach to
  Distributed Artificial Intelligence, edited by
  Gerhad Weiss. MIT Press. 1999, pp27-77
• Anand S. Rao and Michael P. Georgeff. Modeling
  rational agents within a BDI-architecture. In
  Proceedings of the Second International
  Conference on Principles of Knowledge
  Representation and Reasoning, KR '91, 473-484,
  Cambridge, MA, 1991.
• M. Winikoff, L. Padgham, and J. Harland.
  Simplifying the Development of Intelligent
  Agents. In Proceedings of the 14th Australian
  Joint Conference on Artificial Intelligence
  (AI'01), Adelaide, 2001.
• J. P. Müller. The design of Intelligent Agents.
  Springer Verlag. 1996. pp. 7-44.
• FIPA specification. Agent Communication
  Language. http//www.fipa.org/specs/fipa00003/OC0
  0003A.html
• http//fipa-os.sourceforge.net/