Back to first pageBack to first page Centre for Artificial Intelligence of UNL
Browse our site
You are here:

Publication details

Go back
Publication details
Main information
Building Fuzzy Thematic Clusters and Mapping Them to Higher Ranks in a Taxonomy
September 2010
BFTCMHRT
A method in shown to analyze organization activities, say lab research, mapping them to a related hierarchical taxonomy eg. ACM Classification of Computer Subjects. Collected activities data of organization components are presented as fuzzy membership profiles over taxonomy subjects. Profiles generalize in 2 steps: finding fuzzy clusters of taxonomy subjects in the organization; then clusters are parsimoniously mapped to higher ranks. Each innovative step is formalized and solved. Fuzzy clusters of taxonomy leaves are built according to similarity of individual profiles by additive spectral, fuzzy clustering method with model-based stopping conditions not in other methods. As found clusters might not be taxonomy consistent each forms a query set. To lift a query set to higher ranks we use a recursive algorithm minimizing a penalty function for “head subjects” of higher ranks and their “gaps” and “offshoots”. Application to real-world data is made.
Journal
Boris Mirkin, Susana Nascimento, Trevor Fenner, Luís Moniz Pereira
International Journal of Software and Informatics
ISCAS-IFCoLog-Science Press
http://www.ijsi.org/IJSI/ch/reader/issue_list.aspx?year_id=2010&quarter_id=3
4 (special issue of KSEM2010)
3
257-275
-
1673-7288
special issue of select extended papers from KSEM 2010
http://centria.di.fct.unl.pt/~lmp/publications/online-papers/KSEM51_IJSI_Rev.pdf
Export formats
Boris Mirkin and Susana Nascimento and Trevor Fenner and Luís Moniz Pereira, Building Fuzzy Thematic Clusters and Mapping Them to Higher Ranks in a Taxonomy, International Journal of Software and Informatics, Vol. 4 (special issue of KSEM2010), No. 3, Pag. 257-275, ISCAS-IFCoLog-Science Press, http://www.ijsi.org/IJSI/ch/reader/issue_list.aspx?year_id=2010_amp_quarter_id=3, ISSN 1673-7288, <i>special issue of select extended papers from KSEM 2010</i>, (http://centria.di.fct.unl.pt/~lmp/publications/online-papers/KSEM51_IJSI_Rev.pdf), September 2010.
<b>Boris Mirkin, <a href="/people/members/view.php?code=4d69262d034cb8174d039bea8d970836" class="author">Susana Nascimento</a>, Trevor Fenner and <a href="/people/members/view.php?code=6175f826202ff877fba2ad77784cb9cb" class="author">Luís Moniz Pereira</a></b>, <u>Building Fuzzy Thematic Clusters and Mapping Them to Higher Ranks in a Taxonomy</u>, International Journal of Software and Informatics, Vol. 4 (special issue of KSEM2010), No. 3, Pag. 257-275, ISCAS-IFCoLog-Science Press, http://www.ijsi.org/IJSI/ch/reader/issue_list.aspx?year_id=2010_amp_quarter_id=3, ISSN 1673-7288, <i>special issue of select extended papers from KSEM 2010</i>, (<a href="http://centria.di.fct.unl.pt/~lmp/publications/online-papers/KSEM51_IJSI_Rev.pdf" target="_blank">url</a>), September 2010.
@article {BFTCMHRT, author = {Boris Mirkin and Susana Nascimento and Trevor Fenner and Lu\'{\i}s Moniz Pereira}, title = {Building Fuzzy Thematic Clusters and Mapping Them to Higher Ranks in a Taxonomy}, journal = {International Journal of Software and Informatics}, publisher = {ISCAS-IFCoLog-Science Press}, address = {http://www.ijsi.org/IJSI/ch/reader/issue_list.aspx?year_id=2010_amp_quarter_id=3}, volume = {4 (special issue of KSEM2010)}, number = {3}, pages = {257-275}, issn = {1673-7288}, note = {special issue of select extended papers from KSEM 2010}, url = {http://centria.di.fct.unl.pt/~lmp/publications/online-papers/KSEM51_IJSI_Rev.pdf}, abstract = {A method in shown to analyze organization activities, say lab research, mapping them to a related hierarchical taxonomy eg. ACM Classification of Computer Subjects. Collected activities data of organization components are presented as fuzzy membership profiles over taxonomy subjects. Profiles generalize in 2 steps: finding fuzzy clusters of taxonomy subjects in the organization; then clusters are parsimoniously mapped to higher ranks. Each innovative step is formalized and solved. Fuzzy clusters of taxonomy leaves are built according to similarity of individual profiles by additive spectral, fuzzy clustering method with model-based stopping conditions not in other methods. As found clusters might not be taxonomy consistent each forms a query set. To lift a query set to higher ranks we use a recursive algorithm minimizing a penalty function for “head subjects” of higher ranks and their “gaps” and “offshoots”. Application to real-world data is made.}, keywords = {Additive clustering, spectral clustering, fuzzy clustering, parsimonious lift, research activity structure.}, month = {September}, year = {2010}, }
Publication's urls
/publications/view.php?code=fd2ff5c7e069f3af1bbbb5d84ace92c3
/publications/view.php?code=BFTCMHRT
Go back
Centre for Artificial Intelligence of UNL
Departamento de Informática, FCT/UNL
Quinta da Torre 2829-516 CAPARICA - Portugal
Tel. (+351) 21 294 8536 FAX (+351) 21 294 8541

Fundacao_FCT