WEB SITE LIFE CYCLE

A web site in many ways resembles other types of corporate information

systems. Each web site has a limited life span, similar to the water fall

software life cycle model. One major difference is the emphasis on content

development in multimedia applications. The phases of web site

development are as follows: idea formulation, general web site design,

detailed design of web site, testing of an implementation and maintenance.

(1) Idea Formation: During the idea formulation phase, specific targetmarketing

program, content goals and objectives must be set. Since a web

site development project can become very time consuming and a major

capital investment to owners of small businesses, it may be more effective to

identify opportunity of specialized markets big companies have ignored.

Furthermore, the profile of netizens must be carefully studied [Choi99] to

find out who is surfing the net and what these people are looking at. Small

businesses should be aware of the dynamics of the on-line market place and

develop strategies and plan accordingly. The ideas of this phase can pave the

foundation for developing a comprehensive plan for web site design.

(2) Web Site Design: Web site should be integrated into the company's

backbone information system so that the web site can grow along with the

business. To be successful, companies must integrate e-commerce into their

overall business strategies and processes. Moreover, content needs to be

targeted to specific user's needs. Visitor's information should be collected so

that the company will be able to tailor the web pages to the specific needs of

the interested customers. Furthermore, it is important that the web site can be

surfed fast and efficiently. In addition, the users should be involved by

providing an opportunity for them to input suggestions and complaints. The

development of navigational cues and the user interface is of critical

importance. The actual design tasks can be out-sourced for a small company.

Also a new web site should be linked to as many search engines as possible.

This can increase the chance that the web site is visited. Financial

infrastructure should be developed properly as well.

(3) Testing: Once the implementation is complete, the company should

conduct a pilot to test its integrity and effectiveness. The pilot provides an

opportunity to obtain feedback from functional groups, customers and

business partners. It ensures the quality and usability of the site.

(4) Maintenance: It is essential that new content is developed and the web

site is kept refreshed. Timeliness is the key on the web. Moreover,

appointing a web master to manage the site on a day-to-day basis is

imperative. Web master can trouble-shoot any error such as a link to a

defunct web address, track the traffic of the web site, use reader feed back to

build a loyal following and ensure server maintenance and security. Also,

this person or persons should make sure that the company's web site supports

the latest versions of popular browsers.

HIGH PRESENCE AND HIGH TOUCH

Internet and multimedia are changing the rules of the economy and

redefining our businesses and our lives. It is destroying solutions such as

mass production, segmented pricing, and time and distance for big

businesses. A company can develop a web page and advertising campaign

and quickly compete in the world market. This has led to the flattening of the

economy, whereby established companies and individuals doing business on

their own can compete on an equal plane. The small companies that succeed

in challenging the large companies are the ones who can maintain a global

presence and yet make people feel that they are personal and easy to deal

with.

(1) Small companies can interact closely with their customers, so that the

customers feel that they are able to communicate to the small company what

they need, as opposed to the customers merely accepting the mass-produced

product that large companies will sell and not give much ground for

derivation from the product.

(2) Web changed from just a means of advertising, to a medium to

rapidly exchange ideas with potential customers. Since the small company

listens to what they say, it not only results in having a satisfied (and probably

a faithful) customer but increases sales significantly with time.

(3) The Internet's primary advantage in advertising is not so much in

attracting attention and conveying a brief message (the tasks assigned to

traditional advertising media), but lies instead in delivering in-depth, detailed

information. Its real power is the ability to provide almost infinite layers of

detail about a product or service, interactively, at the behest of the user.

However, small companies have to work smarter and respond more

quickly [Murr98]. They have to avoid mistakes and make the best of

possible use to everything. Corporations with big budgets can afford to lose

their investments, while a small company looks at web as survival, not as an

investment.

The small businesses also need to realize having a web site does not

automatically mean that the company will reach millions of potential

customers. It simply means that there is the potential to reach millions of

potential customers. Company has to promote the site through

advertisements, e-mail, links to other sites, and cutting edge multimedia

technology to attract lots of visitors. For a new start-up small company, a

brand new idea is always crucial. Second, multimedia technology should be

used to provide various kinds of services on the web site. Third, once the site

starts catching on and e-mails start rolling in, more and more person hours

should be put into keep up with it all.

Metadata consumption at the proxy

Multimedia adaptation is a key technology for

assuring the quality of end-to-end delivery in the

network. Adaptation can dynamically react on

different kinds of presentation devices and on

unpredictable resource availabilities. Proxy

servers situated in the middle of the delivery

chain constitute an ideal place to control multimedia

delivery and adapt the stream if resource

availability changes. In this context, we built the

Quality-Based Intelligent Proxy (QBIX), which is

a terminal capabilities- and metadata-aware

mapped (RTSP; http://www.rtsp.org) proxy server

that supports real-time adaptation in the compressed

(temporal adaptation only) and

uncompressed domains. Adaptation improves the

hit rate in the proxy cache and lets the proxy act

as a media gateway supporting transcoding in real

time based on metadata.9 One part of the metadata

is sent by the server and describes the video

variations (MPEG-7 VariationSet descriptions),

and the other part is provided by the terminal end

user (in the form of Composite Capabilities/Preferences

Profiles, which we describe later in this article) sending terminal capabilities to the proxy.

The proxy extracts the terminal capability

information from the video request and checks if

it has this music video in its cache in a quality

that matches the terminal capabilities. If the

video is already available, but its properties aren’t

in accordance with the terminal’s characteristics—

for example, the video’s bit rate is too high

to be consumed at the end user’s site—the MPEG-

7 descriptions that accompany the video are

examined by the proxy. They contain hints on

which variations of the original video should be

selected or created (if one doesn’t exist) to meet

the delivery and presentation constraints. These

hints contain the video’s expected size, bit rate,

and quality. The proxy then chooses from among

these variations the one with the highest quality

that meets the restrictions of the end user’s terminal.

The proxy can either load the required

variation from the server or generate it with a

sequence of appropriate transcoding procedures.9

We implemented metadata consumption at the

proxy and the terminal using our Video ToolKit

(ViTooKi). ViTooKi is principally a set of libraries

that support adaptive standard-compliant video

streaming, transcoding, and proxy caching. It supports

MPEG-1/2/4/7/21 and MP3/AAC file types,

stored in various containers like mp4 and avi using

standard protocols with retransmission and the

server, proxy, and player included.

Multimedia Data Cartridge

The MPEG-7 MDC is at the center of the metadata

life cycle because it must manage all the metadata

produced and deliver it to the consuming

elements. The MDC is an extension of an Oracle

database that can store, index, and query multimedia

metadata based on the MPEG-7 standard. It

currently consists of three main parts (see Figure 3,

next page). The core system consists of a multimedia

database schema based on MPEG-7, the Multimedia

Indexing Framework (MIF) supporting

query optimization, and a set of internal and external

libraries for incoming requests and queries.

The MDC has been implemented by a small

group of database programmers with experience

working with the Oracle DBMS kernel. They kept

the extensions to the Oracle database as modular

as possible. Part of these modules, mainly the

indexing framework, is in preparation of a

SourceForge project (http://sourceforge.net/

index.php), which CODAC plans to make available

to the public in Spring 2005.

MPEG-7-based database schema

The multimedia schema relies on the MPEG-7

standard to provide a complete multimedia metadata

schema for low-level descriptions (such as color, texture, and shape for images) and highlevel

descriptions (such as structural and semantic

descriptions) for all media types. We have mapped

the MPEG-7 descriptors, formulated as XML-types,

to object-relational tables to enable fine-grained

querying on these data. (A detailed explanation of

the mapping is available elsewhere.7)

Library support

A set of internal and external libraries are used

for incoming requests and queries. The set of

internal libraries is used as access points to the

core system and consists of InitLib, used for creating

new instances of the MDC data type;

InsertLib, which provides insert functionality of

MPEG-7 documents; DeleteLib, for deleting

MPEG-7 documents; UpdateLib, for updating

parts of stored MPEG-7 documents; and QueryLib,

for query services. Furthermore, external libraries

are used to offer application-specific services.

The services we described in the use case scenario

are VideoLib, for obtaining videos with

semantic search criteria, and AudioLib, for querying

with the humming functionality. Both external

libraries (VideoLib and AudioLib) rely on the

search functionality of the QueryLib, which is

basically a translation of search criteria to complex

SQL and XPATH statements on the schema tables.

Video variation and metadata for content

adaptation

The annotation framework automatically produces

metadata for video variation as a means for

adaptation. In video variation, the hope is to

generate new videos (variation videos) from the

source video, with a reduced data size and quality

by applying a variation or reduction method.

The supported variations include

temporal variation, which reduces the visual

data’s frame rate through frame dropping;

spatial variation, which encodes fewer pixels

(pixel subsampling) and thereby reduces the

level of detail in the frame; and

color variation, which reduces each pixel’s

color depth.

The CODAC project has defined and implemented

two new methods of video variations:5

Object-based variation extracts the foreground

and background objects in a source video and

re-encodes the video with two visual objects.

This facilitates object-based adaptation, which

otherwise would be impossible for a video

encoded without objects. For instance, in our

use case scenario, object-based variation

would be useful for segments with a dynamic

singer foreground and static image background.

This lets an adaptation engine discard,

in case of resource bottlenecks, the static

background image.

❚ Segment-based variation lets us apply variation

methods selectively for video segments based

on the physical characteristics of motion,

texture, and color, thereby minimizing the

quality loss and/or maximizing the reduction

in data size. We segment the source video into

its component shots and automatically select

and apply a set of appropriate methods by

analyzing the degree of physical characteristics

within the segment.

We accomplish variation creation by implementing

a server module called the Variation Factory,

which generates the variations.5 All the

variation methods—including the newly defined

object and segment-based variation methods—are

implemented in the Variation Factory. A source

video can be subjected to a single method or a

combination of them as deemed necessary—for

instance, temporal variation followed by spatial

variation, and so on. At each step, the Variation

Factory produces a variation video, thereby creating

a tree of variations. The Variation Factory is

an application programming interface (API)

including user interfaces to guide the metadata

generation process. The user can control the

adaptation results and rearrange parameters—for

instance, if the perceived quality is too low.

The Variation Factory’s architecture is developed

in Java under Java Media Framework (JMF).

The API supports the integration of audio and

video playback into Java applications and applets.6

The input is video, and the outputs are one or

more adapted variation videos and an MPEG-7

metadata document that describes the source and

the variation videos using descriptors of the VariationSet

description scheme. The MPEG-7 Document

Processor produces metadata. It uses Java

API for XML processing (JAXP), which is generally

used to parse and transform XML documents.

First, a Document Object Model tree is constructed.

Then, the DOM tree is parsed and an

XML text file is produced. The descriptors

include information on the variations’ fidelity,

data size, and priority. We store the created variation

videos in the media server along with the

source video and the MPEG-7 document in the

metadatabase. During delivery, the MPEG-7

metadata document is streamed together with

the requested video.

The Life Cycle of Multimedia

Metadata

During its lifetime, multimedia content

undergoes different stages or cycles from

production to consumption. Content is created,

processed or modified in a postproduction stage,

delivered to users, and finally, consumed. Metadata,

or descriptive data about the multimedia

content, pass through similar stages but with different

time lines.1 Metadata may be produced,

modified, and consumed by all actors involved

in the content production-consumption chain

(see the “Life-Cycle Spaces” sidebar for more

information). At each step of the chain, different

kinds of metadata may be produced by highly

different methods and of substantially different

semantic value.

Different metadata let us tie the different multimedia

processes in a life cycle together. However,

to employ these metadata, they must be appropriately

generated. The CODAC Project, led by

Harald Kosch, implements different multimedia

processes and ties them together in the life cycle.

CODAC uses distributed systems to implement

multimedia processes. Figure 1 gives the architectural

overview of this system.

The project’s core component is a Multimedia

Database Management System (MMDBMS),2

which stores content and MPEG-7-based metadata.

3 It communicates with a streaming server

for data delivery. The database is realized in the

Multimedia Data Cartridge (MDC)—which is an

extension of the Oracle Database Management

System—to handle multimedia content and

MPEG-7 metadata.

Use case scenario

To demonstrate metadata’s life cycle, let’s consider

a use case scenario. A user watches an interesting

music video on a colleague’s screen and

wants to retrieve the same music video. The only

information she retained was that the singer in

the music video was Avril Lavigne, and she

remembers the song’s melody. She can’t ask her

colleague directly, so she wants to access it from

a multimedia database.

The query service of our Multimedia Data Cartridge

(MDC) offers a solution for finding such a

music video. First, the user can enter a query by

thematic means, thus specifying that the music

video’s singer is Avril Lavigne. (See Figure 2a, p. 80,

for the query interface.) In response to the first

query, the service returns several music videos that

pertain to the singer. To narrow the search, the

user can hum the melody to the query with a

humming service, as Figure 2b shows.

The multimedia database retrieves information

on the music video that meets the query request

and delivers it to the user. Such information

includes the full title, full information on the

singer, the production date, and so on, and finally

the address of the media server where the user

can obtain the video. Thus, the user finds out that

the song she has searched for is entitled “Skater

Boy” and can now access the music video.

Luckily, the user is registered to the media

server storing the video and can request the

music video from this server. In addition, the

user specifies in the video request her mobile

device’s terminal capabilities. Unfortunately, the

media server has only copies of the music video

in a quality that doesn’t meet the terminal constraints.

The server examines metadata generated

in postprocessing of the video to generate a

variation of the video with the best possible quality

satisfying the constraints and then delivers

this variation.

Let’s further assume that the request goes over

an authorized proxy cache that examines if a

cached copy is present. If it’s there, but not in the

appropriate quality, delivery-related metadata

describing the possibility for the video to be

adapted to resource constraints is used by the

proxy cache to adapt the video accordingly.

How sound is converted into digital data

When analog information (such as from a CD or live recording) is converted into

digital data, the information is captured and measured into Hertz. The more Hertz you

can capture, the more accurate your recording will be. For example, a sampling rate of

44 kHz (kiloHertz) is far more detailed than a sampling rate of 11 kHz. It is easy to

compare this to a digital picture. The higher the scanning quality, the more detailed the

picture will be. Of course, file size will also increase with higher sampling rates. When

preparing audio for multimedia output, especially web output, it is unnecessary to go

beyond a 22 kHz sampling rate. The reason for this is that this is the average frequency

that is audible to the human ear. Some highs and lows will be lost; however, it is difficult

to detect this, especially with the quality of speakers found on most computers today.

Another technicality to be concerned about when sampling sounds is bit depth.

Bit depth is essentially the dynamic range (or amplitude) of a piece. Bit rate also controls

the resolution of the sound wave (higher bit resolution results in a smoother wave). For

example, a highly structured classical piece would have a high bit depth, because of the

great dynamic differences between a solo flute section and a full orchestral section.

Many pop tunes have less of a bit depth, because they are composed on a more equal

dynamic range. When sampling music for web format, a 16-bit rate is adequate. For

many instances, 8-bit rate may work as well and decrease the file size.

Multi-track recording software

When recording and mixing audio, perhaps the first thing to consider is whether

to go with a Digital Audio Workstation (DAW) or strictly computer software. Digital

Audio Workstations are devices sold as hardware that attach to your computer, allowing

you to store and mix audio through an outside device. Some common components are

mixing boards, multi-track recording devices, and CD burners. The advantages to this is

that processing is usually faster, allowing for more storage space of digital information.

However, the price of DAWs are usually higher than the software versions. Some

common manufactures of DAWs are Tascam and Mackie Designs. Perhaps the top

professional multi-track recording tool is Digidesign's Pro Tools. Digidesign's previous

price of $8,000 made it impossible for private studios and desktop/home studio recorders

to own and operate the equipment. However, this has all changed with the release of

Digi001, which is a simpler version including both the software and hardware interface

for under $1,000. The hardware component is a single-space box and a PCI card that

works with both Mac and PC. A downloadable pro tools software version is also

available for free. Some of the clients who use Digidesign Pro Tools include Nine Inch

Nails, Björk, Smash Mouth, Philip Glass, Third Eye Blind, Paramount Pictures, Canadian

Broadcast Corporation, and such movie projects as Nutty Professor II, The Perfect Storm,

The Matrix, American Beauty, and The Prince of Egypt.

Multi-track software packages are perhaps the most popular, varying in price and

editing power. Some common programs include Samplitude, which samples at 24 bits

and 96 kHz (Windows $69-$399), Vegas™ Audio 2.0 (Windows, $449) which has

unlimited tracks,18 effects, and output options such as streaming media files (WMA and

RM format). Cakewalk® Pro Suite™ is also a popular program (Windows $429-$599).

Cakewalk differs from other programs in that it also allows for MIDI recording and

editing as well as multi-track recording and editing.