17th International Conference on Database and Expert Systems Applications (DEXA'06)
DOI: 10.1109/dexa.2006.135
|View full text |Cite
|
Sign up to set email alerts
|

Text Mining of Business News for Forecasting

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
11
0

Publication Types

Select...
3
3

Relationship

0
6

Authors

Journals

citations
Cited by 11 publications
(11 citation statements)
references
References 3 publications
0
11
0
Order By: Relevance
“…To further evaluate the performance, this study is assessed with two others recent hybrid prediction models by using RMSE and Dstat. The two models compared are TEI@I Nonlinear Integration Forecasting model [10] and EMD-FNN-ALNN model [12]. Both of these models also used BPNN as their training tool.…”
Section: Resultsmentioning
confidence: 99%
See 1 more Smart Citation
“…To further evaluate the performance, this study is assessed with two others recent hybrid prediction models by using RMSE and Dstat. The two models compared are TEI@I Nonlinear Integration Forecasting model [10] and EMD-FNN-ALNN model [12]. Both of these models also used BPNN as their training tool.…”
Section: Resultsmentioning
confidence: 99%
“…Research [12] retrieved information from stock market and used text mining to mine the news and analyse the correlations. In this research, text mining is also applied with Google News to retrieve the information.…”
Section: A Hierarchical Conceptual (Hc) Modelmentioning
confidence: 99%
“…If, however, input is non-numeric, we often use transformations to represent the original information by vectors of numeric-valued features. Exemplarily, we would like to mention the case of text mining, where text is transformed into real-valued feature vectors using the TFIDF transformation (Salton and McGill 1986) and the so-called vector space model; see Kroha et al (2006). The same idea has been applied to clustering of images, music, web pages, among others.…”
Section: Dcc: Types Of Granulationmentioning
confidence: 99%
“…In our previous works (Kroha et al, 2006), and , we succesfully used other methods for text classification but there was no possibility to run the classification algorithms in parallel. In (Kroha et al, 2006), we surprisingly achieved the best results using the Naive Bayes classifier. But we found that there are important classification problems that cannot be decided when using the term frequency as the only object feature.…”
Section: Related Workmentioning
confidence: 99%
“…We used the same stock exchange news in German language as in our previous work (Kroha et al, 2006). However, the time interval was extended from 1.11.1999 until 13.7.2007.…”
Section: Data Preparationmentioning
confidence: 99%