Whats An Intelligent Agent

Published Date: 02 Nov 2017

Intelligent agents are major evolution towards solving difficult problem. Intelligent agents empower both sellers & buyers to accomplish e-commerce transactions by enabling precise efficient & comprehensive searches on vast web community & information repository. Just because of user simplicity & thoroughness, intelligent agents enhance user experience & satisfaction. By operating in background in lieu of user intervention, intelligent agents also circumvent problems related to slow internet access & up prohibitively expensive "surf" & data mining time. 

The technology has evolved & internet has matured to point where sophisticated new generation intelligent agents proliferate. This spage examines intelligent agents & their use by buyers & sellers in e-commerce market space. 

What’s an intelligent agent?

Intelligent agents are software applications which have predefined knowledge base & or learning system about user's goals & wishes & through adaptive reasoning, use dis information to execute their user request. The concept is been around from many decades, ranging in functionality from simple macro prescriptive directions to new generation software dat truly exhibits learning & artificial intelligence. 

Intelligent agents are member of bot family—software programs which are operated unattended, usually on Internet. Agents are sometimes referred as bots. Organizations or Individuals use intelligent agents to perform tasks or functions dat otherwise would involve repetition or human interaction. Operating independently on behalf of users, some intelligent agents mimic human behavior & thought processes & able to make decisions, interact & learn with other intelligent agents. Intelligent agents come stationary & mobile varieties, that they can either reside on individual computer systems or travel server to server across the Internet to carry different tasks.

According to Online, agreement among many authors in field of artificial intelligence that true intelligent agent must be proactive, social, adaptable & autonomous. In early 2000s, intelligent agent technology were still evolving & single agent with all four of these traits hadn’t been created. Multi-agent systems, or groups of intelligent agents in each exhibits one or several of four behaviors, were in making. Nevertheless, many evolving forms of intelligent agents exist in exactly everything from search engine to computer help system.

In world of e-commerce, intelligent agents are known shopping bots & are used by consumers to search product & pricing information on Web. Every shopping bot operates differently, depending on business model used by operator. In 1 scenario, shopping bots direct the users to retailers who, by subscribing for fee, are part of closed system Shopping Yahoo is an example of this model. Open systems are more common arrangement & involve agents that include entire Web in their searches.

Shopping bots are very popular with consumers. In Time, International Data Corp. revealed about four million consumers took advantage of this technology in October 2000 . However, they werenot famous with few companies because of their ability to start bidding wars & eat away profits in process.

Intelligent agent is piece of software that acts for user or the other program in relationship of agency. Such "action on behalf of" implies authority to decide which action may be appropriate .Work is done by allowing people to delegate work that could have done to agent software. Agents can, automate repetitive tasks, make u remember things you forgot, intelligently summarize difficult data, learn from u & even make suggestions to you.

Agents vs. Objects

Agents are more autonomous whereas objects are not autonomous.

Agents have flexible behavior, proactive, reactive & social.

Agents have at least 1 thread of control but objects may have more.

Agents vs. Expert systems

Expert systems are not at all coupled to their environment;

Expert systems are not designed for proactive behavior, reactive.

Expert systems d nt consider social ability

Intelligent Software Agents vs. Intelligent Agents in Artificial Intelligence

Intelligent agents are not just software programs but also be human beings, machines & communities of the human beings.

There are 4 types of intelligent software agents:-

Buyer agents known as 'shopping bots', travel around network retrieving information & work very efficiently for commodity products such as electronic components, books, CD, & other one-size-fits-all products. Amazon.com is good example of shopping bot.

User agents or personal agents are intelligent agents which take action on behalf of user & perform following tasks:

-Check e-mail, sort it according to user's order preference, & alert when important emails arrive.

-Play computer games as patrol game areas

-Assemble customized data, news & reports.

-Find information on subject of our choice.

-Fill out forms on Web automatically storing information for further reference

-Scan Web pages looking for & highlighting text

-Profile synchronization across heterogeneous social networks.

Surveillance Agents & Monitoring r used to observe & report on usually computer systems, equipments. The agents may observe competitors' prices, keep track of companies inventory levels & relay them back to company, watch stock manipulation through insider trading and rumors, etc.

Data mining agent operates in data warehouse discovering information. Data warehouse gets together information from different sources. "Data mining" is process of looking through data warehouse to find information needed to take action.

Intelligent agents can be classified into five classes based on the degree of perceived intelligence & capability.

A simple reflex agent acts on basis of current percept. Agent function is based on condition-action rule. This agent function succeeds when environment is fully observable. Some reflex agents can contain information on current state which allows them to disregard conditions whose actuators are triggered.

Model-based reflex agents handle partly observable environments. Its current state is stored inside agent maintaining some kind of structure describing part of world which cannot be seen. This kind of behavior requires information on how world behaves & works. This kind of additional information completes "World View" model. He keeps track of current state of world using internal model. It chooses an action in same way as reflex agent.

Goal-based agents are models which store information of situations that are desirable. This allows agent a way to choose among many possibilities, selecting one which reaches a goal state.

Utility-based agents only differentiate between goal & non-goal states. It’s possible to define measure of how desirable particular state is. This measure can also be obtained through use of utility function which maps state to measure the utility of the state.

Learning agents: Learning has plus point that it allows agents to initially operate in unknown environments & to become more competent than the initial knowledge alone might allow.

Some of sub-agents that may be a part of an Intelligent Agent :

Decision Agents –are for decision making

Input Agents- process & make sense of sensor inputs

Processing Agents- solve problem such as speech recognition

Spatial Agents- relate to physical real world

World Agents- incorporate combination of the other classes of agents

Believable agents - Agent exhibiting personality through artificial character Physical Agents - percept through sensors & act via actuators.

Temporal Agents - Offer instructions or data acts to computer program or human being & takes program input.

CHARACTERISTICS OF INTELLIGENT AGENTS

Accommodate new problem solving rules incrementally

Adapt online & in real time

Be able to analyze itself in terms of error, success and behavior.

Learn & improve through interaction with environment

Learn quickly from big amount of data

Have memory-based exemplar storage & retrieval capacities

Have parameters to represent long & short term memory.

VALUE OF INTELLIGENT AGENTS IN NETWORKED WORLD

A major value of employing software agents with internet, extranet, & intranet applications is that they are able to locating and filtering data. Time is saved by them making decisions about what is relevant to the user. They are able to sort through network & various databases effortlessly & with unswerving attention to detail in order to get best data. Agents are artificial intelligence's answer to need created by internet-working computers. The main reasons for growth of this technology:

Mundane personal activity: In a fast-speed society, time-bounded people need new ways to minimize time spent on routine personal tasks & devote more time towards professional activities.

Search and retrieval: It is not possible to directly change a distributed database system containing millions data objects. Users relegate task to those agents who will perform tedious, time-consuming, & repetitive tasks of searching databases, retrieving & filtering information, & delivering it back to user.

Repetitive office work: There is need to automate tasks performed by clerical personnel & administrative in functions like customer support or sales in order to reduce labour costs & increase office productivity.

Decision support: There is need for increased support for task performed by good workers who should take timely & knowledgeable decisions to increase effectiveness.

Domain experts: It is recommended to model costly expertise & make it widely available. Expert software agents can model real-world agents like translators, lawyers, stockbrokers, union negotiators, diplomats.

List of tasks performed by the agents includes summarizing, teaching, translating ,watching, empowering, explaining, filtering, guiding, identifying, matching, monitoring, navigating, negotiating, organizing, presenting, querying, reminding, browsing, critiquing, distributing, enlisting reporting, retrieving, scheduling, searching, securing, soliciting, sorting, storing, suggesting, advising, alerting, broadcasting, & so on.

Software agents make networked world less forbidding, save time by reducing effort required to locate & retrieve data & improve productivity by off-loading variety of mindless tasks, mundane, tedious.

ESSENTIALS OF INTELLIGENT AGENTS

There are many possible abilities or traits that people think of when discuss intelligent agents. Four of these traits—temporal continuity, reactivity, autonomy and goal driven— essential to differentiate agents from other types of programs, systems or software objects. Software agents possessing only traits which are often labeled weak or simple. Besides essential traits, software agent may also have additional traits like adaptability, sociability, personality & mobility.

Autonomy: Regular computer programs respond only for directing manipulation. An intelligent agent senses its environment & acts autonomously on it. He can initiate communication, perform tasks & monitor events without direct intervention of humans or others.

Temporal Continuity: Intelligent agent is program to which a user gives a goal or task. Idea is that once a goal or task has been delegated, it is up to agent to work in pursuit of that goal. Agent continues to run—active in foreground or sleep inbackground—monitoring system events that trigger actions.

Reactivity: Intelligent agent responds in a timely fashion to amend in its environment. This characteristic is very crucial for automation and delegation.

Goal Driven: Intelligent agent does a lot more than just simply respond to the changes in its environment. It can also accept high-level requests telling the goals of human user & decide how & where to satisfy requests.

IMPACTS OF INTELLIGENT AGENTS

Intelligent agents are those innovative technologies which offer various benefits to their end users by repetitive tasks or automating complex tasks. There are many potential organizational & cultural impacts of this kind of technology that should be considered. Organizational impacts include transformation of entire operational encumbrance, electronic commerce sector & security overload. Software agents are able to swiftly search the Internet, identify best offers available online,& present this kind of information to end users in aggregate form. Users do not need to manually browse various websites; they are able to locate best deal in matter of seconds. This increases price-based competition & transforms entire electronic commerce sector into uniform perfect competition market. Cultural effects of implementation of software agents include privacy attrition, social detachment, trust affliction & skills erosion. Few users may not feel comfortable fully delegating most important tasks to software applications. To act on a user’s behalf, a software agent must have complete understanding of user’s profile, including his personal preferences. This may lead to unpredictable privacy issues. When users start relying on software agents for communication activities, they can lose contact with other human users & look at word with eyes of their agents.

Intelligent agents are member of bot family—software programs that operate usually on the Internet and unattended. Agents are sometimes referred to as bots. Organizations or individuals use intelligent agents to perform tasks or functions that otherwise involve human repetition or interaction. Operating independently on behalf of users, some intelligent agents mimic human behavior & thought processes & are able to make decisions, interact and learn with other intelligent agents. Intelligent agents come in stationary & mobile varieties, meaning that they can either reside on individual computer systems or travel server to server across Internet to carry out tasks.

According to Online, there is an agreement among several authors in field of artificial intelligence that true intelligent agent must be adaptable, social, autonomous and proactive. In early 2000s, intelligent agent technology was still evolving & single agent with all 4 of these traits had not been created. Groups of intelligent agents or Multi agent systems in which each exhibits single or several of four behaviors, were in development. Many evolving forms of intelligent agents exists in every from search engines to computer help systems.