Další formáty:
BibTeX
LaTeX
RIS
@inproceedings{568397, author = {Hroza, Jiří and Žižka, Jan}, address = {Germany}, booktitle = {Computational linguistics and Intelligent Text Processing}, keywords = {machine learning; text categorization; text filtration; text similarity; k-NN; ranking}, language = {eng}, location = {Germany}, isbn = {3-540-24523-5}, pages = {608-611}, publisher = {Springer-Verlag Berlin Heidelberg}, title = {Selecting Interesting Articles Using Their Similarity Based Only on Positive Examples}, year = {2005} }
TY - JOUR ID - 568397 AU - Hroza, Jiří - Žižka, Jan PY - 2005 TI - Selecting Interesting Articles Using Their Similarity Based Only on Positive Examples PB - Springer-Verlag Berlin Heidelberg CY - Germany SN - 3540245235 KW - machine learning KW - text categorization KW - text filtration KW - text similarity KW - k-NN KW - ranking N2 - The task of automated searching for interesting text documents frequently suffers from a~very poor balance among documents representing both positive and negative examples or from one completely missing class. This paper suggests the ranking approach based on the k-NN algorithm adapted for determining the similarity degree of new documents just to the representative positive collection. From the viewpoint of the precision-recall relation, a~user can decide in advance how many and how similar articles should be released through a filter. ER -
HROZA, Jiří a Jan ŽIŽKA. Selecting Interesting Articles Using Their Similarity Based Only on Positive Examples. In \textit{Computational linguistics and Intelligent Text Processing}. Germany: Springer-Verlag Berlin Heidelberg, 2005, s.~608-611. ISBN~3-540-24523-5.
|