MURDOCH RESEARCH REPOSITORY
http://dx.doi.org/10.1049/cp.2009.1852
Fung, C.C., Thanadechteemapat, W. and Wong, K.P. (2009)
Summarizing information from Web sites on distributed power
generation and alternative energy development. In: 8th IET
International Conference on Advances in Power System Control,
Operation and Management (APSCOM 2009), 8 - 11 November,
Hong Kong.
http://researchrepository.murdoch.edu.au/13169/
Copyright © 2009 IEEE
Personal use of this material is permitted. However, permission to reprint/republish
this material for advertising or promotional purposes or for creating new collective
works for resale or redistribution to servers or lists, or to reuse any copyrighted
component of this work in other works must be obtained from the IEEE.
SUMMARIZING INFORMATION FROM WEB SITES ON DISTRIBUTED POWER
GENERATION AND ALTERNATIVE ENERGY DEVELOPMENT
Chun Che Fung
School of Information
Technology,
Murdoch University,
WA 6150, Australia
Wigrai Thanadechteemapat
School of Information
Technology,
Murdoch University,
WA 6150, Australia
Kit Po Wong
Department of Electrical
Engineering, The Hong Kong
Polytechnic University,
Hung Hom, Hong Kong
ABSTRACT
The World Wide Web (WWW) has become a huge
repository of information and knowledge, and an
essential channel for information exchange. Many
sites and thousands of pages of information on
distributed power generation and alternate energy
development are being added or modified constantly
and the task of finding the most appropriate
information is getting difficult. While search engines
are capable to return a collection of links according
to key terms and some forms of ranking mechanism,
it is still necessary to access the Web page and
navigate through the site in order to find the
information. This paper proposes an interactive
summarization framework called iWISE to facilitate
the process by providing a summary of the
information on the Web site. The proposed approach
makes use of graphical visualization, tag clouds and
text summarization. A number of cases are presented
and compared in this paper with a discussion on
future work.
1. INTRODUCTION
The world is facing an urgent challenge of how to
develop alternative energy and distributed power
generation efficiently. Most of the world’s energy
demands at present are still relying on crude oil, coal
and natural gas [1]. It can be said that such resources
are finite and there are limitations on the supply side
of the equation and eventually such resources will
not able to meet the upward demands. In 2006, it has
been reported that the total world production of
energy was already unable to maintain further
growth in demand [2]. Also, the price of crude oil
has undergone extreme fluctuations [3]. For
example, the oil prices in 2008 had been rapidly
increased in the first half and then dropped steeply in
the second half of the year [4]. This has a significant
impact on the world economy [5]. On the other
hand, environmental issues due to ‘greenhouse
effects’ as a consequence of CO
2
emission from
fossil fuel consumption, may have contributed
towards the global climate change [6]. That will
affect many generations to come and it is predictable
that this will have enormous social, economic and
environmental implications [8]. Hence, there is an
urgent need to develop alternate energy and
distributed power generation which has to be
environmental friendly, sustainable, reliable and
hopefully, cheap.
One of the solutions is to reduce the world’s
dependence on fossil fuels by escalating the reliance
on alternative or sustainable energy sources. Many
governments, research and commercial
organizations have already committed considerable
amount of effort and resources to this cause. Such
effort can be further facilitated by providing and
exchanging up-to-date information and knowledge
among the researchers and related bodies in the
discipline. An effective means for the provision and
distribution of information cheaply and quickly is to
use the World-Wide-Web on the Internet. At
present, the number of Web sites has significantly
increased approximately ten times from 1996 to
2009 at 224 million Web sites [9]. Information on
alternative energy have already been provided by
many Web sites as indicated in a survey study by the
authors [10]. However, the enormous amount of
information on the Web leads to the phenomenon of
“information overload” and this has caused much
burden on the researchers [11]. An inefficient
process or inability of locating the required
information may lead to wasted efforts and hinders
the promotion and development of alternative
energy. Although search engines are currently the
de-facto tool that most searches on the web are
based on, a large amount of results are normally
produced due to information overloading. Users still
have to visit each individual Web page in order to
identify whether it contains the desired information.
On the other hand, search engines also deployed
their own algorithms or mechanisms in ranking the
results and some of them may be based on financial
interest. Hence, the results do not provide any
context or quality of the sites. As a solution to these
problems, a new approach termed iWISE, an
intelligent Website Information Summarization
Engine, is proposed in this paper. The proposed
approach presents the information from the web sites
with a comprehensive overview by using
visualization and summarization techniques. This
approach is believed having the potential to reduce
the time to identify and traverse the information in a
Web site.
This paper is organized with an introduction and
aims of this paper. Summarization Technology is
then described in Section 2, and visualization and
summarization engine are explained in Section 3.
Application towards alternative energy information
on the web is illustrated in Section 4, and the
conclusion including discussion on future work in
Section 5 followed by acknowledgement and
references.
2. SUMMARIZATION TECHNOLOGY
A summary is used to provide the key meaning of an
original source of information in a reduced and
concise representation. This representation could be
in the forms of graphics, tables or text. Computer
programs using various intelligent techniques may
also be used to create a summary. A number of the
related methodologies and techniques in iWISE are
explained in the following subsections.
2.1 Text Summarization
Text summarization is a process used to produce a
summary based on the original information. The
summary can be produced by extraction or
abstraction from the original text [12]. A summary
by extraction is to find a part of text that can be
considered as indicative of the content using
choosing the best rank of sentences [11]. The most
common technique is the statistical approach that
examines sentences, phrases, or words based on
frequency or histogram. They are then ranked and
the highest rank sentences and phrases are used to
produce a summary. There are a diversity of such
techniques being developed and applied.
The other approach is summary by abstraction,
which aims to a summary using natural language
processing (NLP) techniques based on the original
semantics. NLP attempts to understand and process
human languages by extracting key phrases of the
original text [13]. It can also be applied with
machine learning using both supervised and
unsupervised learning approaches. For example, the
original text and its key phrases can be set as a
training set in the form of a classifier, which
determines a threshold for choosing the key phrases.
This will set limits on different domains and the
classifier has to be defined with some rules in each
domain with specific knowledge, and a number of
training data sets are needed.
On the other hand, unsupervised learning can also be
used as a part of the summarization process. One of
the techniques is based on an algorithm to examine
the original text in order to identify key phrases,
which represent most of the main meaning in the
original document.
Furthermore, there is a hybrid approach which
applies both statistics and natural language
processing for producing useful summaries [14].
2.2. Tag Cloud
Tag cloud is a group of tags that are extracted from
related words in the original text [15]. It can
represent the original information in a form of an
image of short phrases or words, which is considered
to be a form of visual summary. Tags can be
extracted by applying statistical approaches such as
finding frequency of words. Unsupervised learning
techniques such as clustering can be applied to
identify and select the words that are properly
related to the original document [16].
The cloud normally shows the extracted tags in
different illustration according to style, color and
size from the underlining words or phrases in the
original text. Visualization of the cloud may use
diverse algorithms to fit the words in one limited
layout. The limited area can be fitted the tags by
using spatial clustering algorithm [16] [17]. On the
other hand, the cloud may also be represented in
different ways such as in a circular layout [18].
Further filtering of the tags could be applied based
on specific domain and elimination of non-
representative words.
2.3. Summarization of Web pages
Web pages in electronic format can provide better
interactive and visual information than normal print
based documents. The information on the Web page
can include sound, video, hyperlinks, and
interactivity. Web page uses Hypertext Markup
Language (HTML) to format its content in a normal
text file consists of HTML tags and content. The
latest Web pages may also include the use of
Extended Hypertext Markup Language (XHTML),
Extensible Markup Language (XML) and Script
languages. The formatted contents are required to be
displayed in a browser. If the content is
predominately text, it can be summarized by text
summarization techniques. A summary of the
information on the Web page has more steps than
text summarization due to the inclusion of HTML
tags and other script languages which have to be
removed first. Furthermore, the content also has to
be rearranged before working on the summarization
process. For example, the content that is displayed in
tables needs to be carefully restructured.
On the other hand, metadata in a Web page can be
used as a form of its summary because the metadata
is used metatags to provide a short description.
Normally, the metadata are not displayed and they
can be used to facilitate an automatic indexing
process [19].
3. iWISE - INTELLIGENT WEBSITE
INFORMATION SUMMARIZATION
ENGINE
This paper proposes a novel approach called iWISE,
which integrates summarization techniques with
visualization techniques in order to provide a
summary of a Website. iWISE provides various
aspects of summaries of a Website using tag cloud,
text summarization and Document Type Views
(DTV) as well as a thumbnail of the Web pages
within the site. The engine operates in real time and
it is interactive, whereby a user can navigate the
Web page through this engine. It also allows a form
of cross checking between the techniques to enable
the users to gain a higher degree of confidence on a
summary. For instance, most of words in tag cloud
should match to the key phrases in a summary from
text summarization. Moreover, iWISE can be used
as a tool for comparing Web sites in order to select
the Web pages which deem to provide the most
appropriate information according to needs of the
users.
Text Summarization is a feature aiming to produce
a summary of the Web page, while the summary
may be highlighted sentence lists or paragraphs. The
summarization engine provided from
www.extractorlive.com is used in this paper as an
example of this approach. The home page of
ScottishPower Renewables and the result of text
summarization are shown in Figure 1 and 2,
respectively.
Figure 1. Home page of ScottishPower Renewables
1
1
http://www.scottishpowerrenewables.com
Figure 2. A summary of the home page of the
ScottishPower Renewables
Tag Cloud is another key summarization output in
iWISE. The example used in this paper is based on
results from the TagCrowd website
2
. While it is
recognized that there are no specific meanings on
tag clouds [20], iWISE can address this issue by
linking the related results to text summarization. The
example tag cloud of the ScottishPower Renewables
is shown in Figure 3 below
Figure 3. A tag cloud of the home page of the
ScottishPower Renewables
Document Type View (DTV) is a visual summary
output from the iWISE approach. This is comparable
to the Webtrustmetrix developed by one of the
authors used on the Ainibot Web site
3
. This feature
is used to provide an illustration on the general
structure of a Website.
The DTV result of the ScottishPower Renewables
Website is shown in Figure 4. In addition, a small
thumbnail of the Web page can be displayed
alongside with the overview of the Web page when a
user points to the hyperlink node on the DTV. This
is illustrated in Figure 5. Thus, a user may visit any
Web page that is displayed in the DTV by the “Drill
Down” function. When the linked node has been
drilled down, the engine will repeat the working
process to produce text summary, tag cloud and
Document Type View, which will be synchronized,
at the same time. An example of a drilled down page
is shown in Figure 6.
2
http:// www.tagcrowd.com
3
http://www.ainibot.org/webtrustmetric
Figure 5. A thumbnail on a linked node is shown in
the drill down feature
Figure 6. Illustration of a result after clicking a
linked node
Figure 4. An actual result in this approach