![](bg1.png)
AN EXTRACTION METHOD OF TIME-SERIES NUMERICAL DATA
FROM ENTERPRISE PRESS RELEASES
Masanori Akiyoshi, Mayu Gen, Masaki Samejima, Norihisa Komoda
Osaka University
Yamadaoka 2-1, Suita, Osaka 565-0871, Japan
Keywords:
Time-series numerical data, enterprise press releases, automatic extraction.
Abstract:
This paper addresses an extraction method of time-series numerical data from enterprise press releases for
business strategy design. Business strategy consists of logical actions for continuously producing enterprise
outcome. The business strategy design process that is partially based on competitive environment analysis may
extremely resort to professional skills so far. To enhance and accelerate the competitive environment analysis,
we focus on press releases of competitors in order to extract numerical data related to products or services.
Sentences in press releases are well organized and grammatically correct. Therefore such extraction is simply
done by identifying the keywords of products or services and the unit indicator co-occurrence. In addition to
such simple approach, we clarify the specific rules to applying our method to practical press releases.
1 INTRODUCTION
Recently business is considered to be under unpre-
dictable environment, which means rapid changes
of consumers’ demand cause sudden drop of profit
and weaker positioning in the target market. When
present corporate management is in excellent state,
future corporate management is no longer on the elon-
gation of it. To judge investment adequately, it is in-
dispensable to analyze various factors, for instance,
not only financial aspects but also aspects of organi-
zational growth, customer satisfaction, and so forth.
On the other hand, along with the development of en-
terprise information systems, it becomes possible to
judge business operations rapidly and adequately, for
instance, sales planning by using OLAP (on line an-
alytical processing) of data warehouse or data min-
ing technique. Even if such business operations are
succeeded, continuous growth of business is not at-
tained without business strategy. Business strategy
has become more critical to corporate management
(Loebbecke, 2003).
Business strategy consists of logical actions for
continuously producing enterprise outcome. Such de-
sign may extremely resort to professional skills so
far, because the design process involves repetitive
work of various analysis, goal setting, and investiga-
tion through simulation, which are inherently human-
centric work. It takes much time to design business
strategy only by hand of human professionals. There-
fore there is a strong need to make it possible to de-
sign business strategy by newly emerging information
technology, which is expected to support human pro-
fessionals.
While there are several possible approaches to sup-
port human professionals, we focus on document
analysis-based support for repetitive analysis. Busi-
ness documents such as enterprise annual reports and
press releases indicate applied results of business
strategy. Market-related documents such as consumer
opinion collected through antenna shops or a callcen-
ter include potential needs for products or services.
Therefore we think analysis on business documents
is inevitable to design business strategy (Akiyoshi,
2005).
In this paper, as analysis on business documents,
we focus on press releases of competitors in order
to extract numerical data related to products and ser-
vices, and so forth. We first describe some features
of press releases, and then propose our extraction
method in detail. Finally experimental results are also
discussed.
221
Akiyoshi M., Gen M., Samejima M. and Komoda N. (2006).
AN EXTRACTION METHOD OF TIME-SERIES NUMERICAL DATA FROM ENTERPRISE PRESS RELEASES.
In Proceedings of the First International Conference on Software and Data Technologies, pages 221-225
DOI: 10.5220/0001318202210225
Copyright
c
SciTePress