04.11.2014 Views

elektronická verzia publikácie - FIIT STU - Slovenská technická ...

elektronická verzia publikácie - FIIT STU - Slovenská technická ...

elektronická verzia publikácie - FIIT STU - Slovenská technická ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

246 Selected Studies on Software and Information Systems<br />

word distribution metrics. Then a probabilistic information extraction model is employed to<br />

find contact information and person names in the homepages. Newly extracted people who<br />

are coreferent to already discovered people are determined. Links are placed in the social<br />

network between a discovered person and the owner of the web page on which the person<br />

was discovered. The extraction was done using conditional random fields.<br />

E-mail<br />

Web<br />

Keyword<br />

extraction<br />

Person name<br />

extraction<br />

Name<br />

coreference<br />

Homepage<br />

retrieval<br />

Contact<br />

Information and<br />

person name<br />

Extraction<br />

names<br />

Social Network<br />

Analysis<br />

Figure 8-11. Overview of a system performing social network extraction, according to [20].<br />

The approach suffers from names and web appearance ambiguity problems. In some cases, it<br />

can recursively extract social network of a namesake of people from original social network.<br />

This issue can be fixed as shown in [7].<br />

References<br />

[1] Agichtein, E.: Web Information Extraction and User Modeling: Towards Closing the<br />

Gap. IEEE Data Eng. Bull., 2006, vol. 29, no. 4, pp. 37–44.<br />

[2] Andrejko, A., Barla, M., Bieliková, M.: Chap. Ontology-based User Modeling for Webbased<br />

Information Systems. In: Advances in Information Systems Development. Springer,<br />

2007, pp. 457–468.<br />

[3] Atzenbeck, C., Tzagarakis, M.: Criteria for Social Applications. In Vassileva, J., Tzagarakis,<br />

M., Dimitrova, V., eds.: Socium: Adaptation and Personalisation in Social Systems:<br />

Groups, Teams, Communities. Workshop held at UM 2007, 2007, pp. 45–49.<br />

[4] Barla, M.: Interception of User⁄s Interests on the Web. In Wade, V., Ashman, H.,<br />

Smyth, B., eds.: Adaptive Hypermedia and Adaptive Web-Based Systems, AH’06. LNCS<br />

4018, Dublin, Ireland, Springer, 2006, pp. 435–439.<br />

[5] Barla, M., Andrejko, A., Bieliková, M., Tvarožek, M.: User Characteristics Acquisition<br />

from Logs with Semantics. In Kelemenová, A., Kolář, D., Meduna, A., Zendulka, J.,<br />

eds.: ISIM ´07: Information Systems and Formal Models, Slezská universita v Opavě, 2007,<br />

pp. 103–110.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!