Index Construction for Linear Categorisation
Edit Page
Report
Scan day: 02 March 2014 UTC
-45
Virus safety - good
Description: By Vaughan R. Shanks and Hugh E. Williams, RMIT University, Melbourne, Australia. Proceedings of the twelfth international conference on Information and knowledge management, 2003. A problem with iterative training techniques for automatic text categorisation such as Support Vector Machines (SVM) is that during the learning phase, they require the entire training collection to be held in main-memory, which is infeasible for large training collections such as DMOZ or large news wire feeds. The authors present techniques which permit automatic categorisation using very large training collections, vocabularies, and numbers of categories. ODP is mentioned as a possible set of training data.
%PDF-1.3 3 0 obj > stream x YY“ãÆ ~Ÿ_¡¼QU#šdót^b;Ž½N¶6ÉNìJÖ~ ¨‰Y2+ú÷ùpP¤FtMM±}hàÐò7þüÆ®‰£Mœún›¢~úúåé‹¿˜dã'nàG›—Wžúrøä¼kvëGÎÿ¶» òœoÚ¦º±Ê¶ÊkÛIãoecsm“¼êØveŸóä_^~#|7‹ø“¹Y’nvAꦙϧý¸õ}ßÉÇã)×íÿéÊ÷#(Ÿûí.¤Î÷˜ äouø§mè9eU•yÝÓQ›]Q¢p³ó3×)ïþ±8µm%Ú×I úVùþX”¶)¬tòæ w „¬óYæ—mâ9¶85mÕ¯Û8rž•Û÷ï^¤õ¯¦Ü‘óiÁv}9\™$tCcÀ•q³P¸úîïdÉ×*9“š²iÛ&
Size: 2048 chars
Contact Information
Email: —
Phone&Fax: —
Address: —
Extended: —
WEBSITE Info
| Page title: | |
| Keywords: | |
| Description: | |
| IP-address: | 131.170.24.36 |