Transcript Electronic Commerce

Chapter 5: Architectural requirements of EC
IT357 Electronic Commerce
1
Architectural requirements of EC
1.
2.
3.
4.
5.
Telecommunication infrastructure
Decision support systems
Interoperability
Storage and retrieval of multi media information
Markup languages
27 June 2008
IT 357 - Chapter 5
2
Architectural requirements of EC
•
•
•
•
Scalable
Enables resource sharing
Supports asymmetrical protocols
Provides transparency of data locations and decision
support
• Also supports multi-media information, allow mix and
match of various software and hardware platforms
27 June 2008
IT 357 - Chapter 5
3
Telecommunication infrastructure
•
•
OSI - Open Systems Interconnection: Proposed by the International Standards Organization
The goal - to allow one computer to communicate with another irrespective of its origin or
manufacturer.
27 June 2008
IT 357 - Chapter 5
4
Cont’d: Telecommunication infrastructure
Open Systems Interconnection
1. Physical Layer
– Consists of the transmission media for the bit streams and
its interface with computers, e.g. fiber optic cables
2. Data link layer
– Concerned with information transmission over a single
channel or link.
•
•
•
•
Acknowledgement of transmissions
Error detection
Sequencing
E.g. Ethernet, ATM
– The data link layer is often implemented in software as a
"network card driver".
27 June 2008
IT 357 - Chapter 5
5
Cont’d: Telecommunication infrastructure
Open Systems Interconnection
3. Network Layer
– Enables simultaneous use of multiple links to increase
transmission performance.
•
•
•
Routing
Flow control
End to end ack for multi link paths
•
Host to network interfaces Data link layer
– Responsible for end to end packet delivery
4. Transport Layer
– Network addressing
– Establishment of virtual circuits
– Procedures for entering and departing from the network.
27 June 2008
IT 357 - Chapter 5
6
Cont’d: Telecommunication infrastructure
Open Systems Interconnection
5. Session Layer
– Concerned with establishing communications between
given pair of users
– Starting, stopping and controlling the communication.
6. Presentation Layer
– Display formatting and editing of inputs and outputs
7. Application Layer
– Concerned with the business functions the users are trying
to perform.
27 June 2008
IT 357 - Chapter 5
7
Cont’d: Telecommunication infrastructure
TCP/IP
• Covers the top layers of the OSI excluding the
Physical and Data link layers
ATM
• Real-time transmission of multi-media data
• Interoperable with local, metropolitan and wide area
networks
27 June 2008
IT 357 - Chapter 5
8
Decision support systems
Data warehouses
•
Refers to a collection of data that is
–
–
–
–
–
•
Subject oriented
Integrated
Time varying
Non volatile and
Designed to support faster and better decision making
Data warehouses Vs. Transaction Systems
–
–
–
–
–
27 June 2008
A data warehouse application focuses on data analysis
(Online analytical processing - OLAP)
A traditional database application emphasizes on Online transaction
processing – OLTP
Data warehouse applications require summarized historical data derived
from internal and external sources.
OLTP applications require detailed current data derived primarily from
internal sources.
Data warehouse is several orders of magnitude larger than an operational
data base.
IT 357 - Chapter 5
9
Decision support systems
Developing OLAP applications:
• Multiple sources of data pose difficulties:
– Inconsistency in data representation coding and format
– Erroneous and incomplete data
• Data Cleansing
– Done by data migration tools such as Warehouse Manager that rely
on the user to define the mappings between source and target
data.
• Loading and uploading data
– Refers to the process of populating the data warehouse from
external data sources.
– Preprocessing needed for integrity checking, sorting summarization
aggregation and data partitioning (for parallel loading).
27 June 2008
IT 357 - Chapter 5
10
Decision support systems
Developing OLAP applications
• Refreshing data warehouses
– Refreshing the data warehouse based on the changes to the
operational database is important.
– Can be done by:
• Synchronous replication
– Provides tight consistency between data stores with zero latency
– Expensive and not essential in many applications
• Asynchronous replication
– A table in the warehouse is treated as a remote copy of the table
in the operational database.
– The actual updating is done at a later time .
27 June 2008
IT 357 - Chapter 5
11
Decision support systems
Knowledge Discovery in Databases (KDD)
– Refers to the tools and techniques used to analyze
data in organizational databases to find useful
patterns in data.
Data Mining
• Deals with specific algorithms and methodologies for data
analysis with the intent of finding useful patterns.
• Consists of two tasks:
– Prediction - Identification of unknown values for the variables of
interest
– Description - Analysis of data in humanly interpretable forms.
27 June 2008
IT 357 - Chapter 5
12
Interoperability
• Interoperability provided by:
• CORBA
– Common Object Request Broker Architecture
– Object Management Group’s (OMG) open distributed object
computing infrastructure.
– Allows applications to communicate with one another irrespective
of the location or design.
• Software agents
– Computer programs that employ AI techniques to provide active
assistance to users with computer based tasks.
27 June 2008
IT 357 - Chapter 5
13
Interoperability
How does CORBA ensure interoperability?
• The clients and objects interact using one or more Object
Request brokers (ORB)
• ORB is the middleware that establishes the client server
relationships between objects.
• Using an ORB a client can invoke a method on a server object,
which can be on the same machine or across a network.
• The ORB intercepts the call and is responsible for
–
–
–
–
27 June 2008
finding an object that can implement the request
passing the parameters
invoking its method
returning results
IT 357 - Chapter 5
14
Interoperability
Software Agents
•
Assist users in different ways
•
•
•
•
•
•
Information filtering
Information retrieval
Mail management
Meeting scheduling
Selection of books, music etc.
Example: Maxims - An agent that helps with e-mails.
– Learns to prioritize, delete, forward, sort and archive mail messages on
behalf of the user.
•
•
•
PersonaLogic - PersonaLogic allows users to specify product features
uses a constraint satisfaction algorithm to filter through product space
to retrieve an ordered set of products.
Firefly - uses an automated collaborative filtering method to rate and
recommend products to shoppers.
BargainFinder and Jango
– Takes a product name as input, obtains price information from other web
sites and offers a price comparison.
27 June 2008
IT 357 - Chapter 5
15
Storage and retrieval of multi media information
•
•
Based on collecting content representative metadata - data
about media objects stored
Metadata is classified as:
–
–
–
–
–
•
•
•
Content dependant
Content descriptive
Content independent
The metadata has to have efficient linkages with the physical
location of the information.
The data retrieved should be presented to the users in the
decreasing order of retrieval status value
Objects are stored in digital libraries based on feature similarity.
A feature is a set of characteristics such as text strings in text
documents, colour texture and objects in video frames.
These features can be used to index and retrieve objects
27 June 2008
IT 357 - Chapter 5
16
Storage and retrieval of multi media
information
•
Storage strategy is based on:
– Type of object
– Retrieval requirements of the stored object.
•
Storage strategies can be classified under:
– Single disk storage
– Multiple disk storage
– Multiple disk storage with data striping
27 June 2008
IT 357 - Chapter 5
17
Markup Languages
• Markup - originally used to describe annotations or
other marks within a text to instruct a typist how a
passage should be laid out.
• Markup language - a set of mark up conventions
used for encoding of texts.
• A markup language must specify:
–
–
–
–
27 June 2008
What markups are allowed
What markups are required
How markups are distinguished from text
What markup means
IT 357 - Chapter 5
18
Markup Languages
• SGML
– Describes a generalized mark up scheme for representing
the logical structure of the document in a system and
platform independent manner.
– Uses a Document Type Definition (DTD) model to describe
each element of the document
– Different documents of the same DTD can be processed
uniformly.
– Programs can be written to take advantage of the structure
information and made to behave more intelligently.
27 June 2008
IT 357 - Chapter 5
19
Markup Languages
• HTML
– Invented by Tim Bernes Lee
– The publishing language of the WWW.
– Gives users means to:
•
•
•
•
Publish online documents
Retrieve online information
Design forms for conducting transactions with remote services
Include spread sheets, video clips etc in the documents
– Drawbacks
• Does not support encoding of the structure and meaning of structured
documents -e.g. spread sheets, address books, technical drawings etc.
• Thus such documents cannot be effectively indexed or searched.
• Formal descriptions of content and services are absent - Automated
information extraction from HTML pages is therefore difficult
27 June 2008
IT 357 - Chapter 5
20
Markup Languages
•
XML
– Relatively New
– Combines the merits of HTML and SGML
– Used for designing text formats for structured data such as spread sheets
and address books.
•
XML vs HTML
– Uses tags (<words>)and attributes (name=“value”) just like HTML
– But the interpretation of the tags is left to the application entirely.
•
XML Characteristics
–
–
–
–
–
•
XML files are text files meant to be read by programs or programmers.
Programmers can debug applications more easily in XML.
Unlike HTML errors are not tolerated
XML files are larger than comparable binary formats.
Is license free, platform independent and well supported
XBRL
27 June 2008
IT 357 - Chapter 5
21