% %                 Introduction%                 ============% % In my work on context-sensitive learning, I used the "Deterding Vowel% Recognition Data", but I found it necessary to reformulate the data.% Implicit in the original data is contextual information on the% speaker's gender and identity. For my work, it was necessary to make% this information explicit. The file "vowel-context.data" adds the% speaker's sex and identity as new features. The format of the data file% is described below.% % % Peter Turney% peter@ai.iit.nrc.ca% % % %                 References%                 ==========% % P. Turney. "Robust Classification With Context-Sensitive Features."% Proceedings of the Sixth International Conference on Industrial% and Engineering Applications of Artificial Intelligence and Expert% Systems (IEA/AIE-93): 268-276. 1993.% % URL: ftp://ai.iit.nrc.ca/pub/ksl-papers/NRC-35074.ps.Z% % % P. Turney. "Exploiting Context When Learning to Classify."% Proceedings of the European Conference on Machine Learning% (ECML-93): 402-407. 1993.% % URL: ftp://ai.iit.nrc.ca/pub/ksl-papers/NRC-35058.ps.Z% % % %                 File Structure%                 ==============% % %         Column          Description%         -------------------------------%         0               Train or Test%         1               Speaker Number%         2               Sex%         3               Feature 0%         4               Feature 1%         5               Feature 2%         6               Feature 3%         7               Feature 4%         8               Feature 5%         9               Feature 6%         10              Feature 7%         11              Feature 8%         12              Feature 9%         13              Class% % % % %                 Numerical Codes%                 ===============% % %         Speaker         Code Number%         ---------------------------%         Andrew          0%         Bill            1%         David           2%         Mark            3%         Jo              4%         Kate            5%         Penny           6%         Rose            7%         Mike            8%         Nick            9%         Rich            10%         Tim             11%         Sarah           12%         Sue             13%         Wendy           14% % % %         Set             Number%         ---------------------------%         Train           0%         Test            1% % % %         Sex             Number%         ---------------------------%         Male            0%         Female          1% % % %         Class           Number%         ---------------------------%         hid             0%         hId             1%         hEd             2%         hAd             3%         hYd             4%         had             5%         hOd             6%         hod             7%         hUd             8%         hud             9%         hed             10% % % % % %         Speaker         Code Number     Sex             Train/Test%         ---------------------------------------------------------------%         Andrew          0               0               0%         Bill            1               0               0%         David           2               0               0%         Mark            3               0               0%         Jo              4               1               0%         Kate            5               1               0%         Penny           6               1               0%         Rose            7               1               0%         Mike            8               0               1%         Nick            9               0               1%         Rich            10              0               1%         Tim             11              0               1%         Sarah           12              1               1%         Sue             13              1               1%         Wendy           14              1               1% % % Num Instances:  990% Num Attributes: 14% Num missing:    0 /  0.0%%%     name                      type enum ints real     missing    distinct  (1)%   1 'Train or Test'           Enum 100%   0%   0%     0 /  0%     2 /  0%   0% %   2 'Speaker Number'          Enum   0% 100%   0%     0 /  0%    15 /  2%   0% %   3 'Sex'                     Enum   0% 100%   0%     0 /  0%     2 /  0%   0% %   4 'Feature 0'               Real   0%   0% 100%     0 /  0%   853 / 86%  74% %   5 'Feature 1'               Real   0%   0% 100%     0 /  0%   877 / 89%  78% %   6 'Feature 2'               Real   0%   0% 100%     0 /  0%   815 / 82%  67% %   7 'Feature 3'               Real   0%   0% 100%     0 /  0%   836 / 84%  71% %   8 'Feature 4'               Real   0%   0% 100%     0 /  0%   803 / 81%  66% %   9 'Feature 5'               Real   0%   0% 100%     0 /  0%   798 / 81%  64% %  10 'Feature 6'               Real   0%   0% 100%     0 /  0%   748 / 76%  57% %  11 'Feature 7'               Real   0%   0% 100%     0 /  0%   794 / 80%  64% %  12 'Feature 8'               Real   0%   0% 100%     0 /  0%   788 / 80%  63% %  13 'Feature 9'               Real   0%   0% 100%     0 /  0%   775 / 78%  60% %  14 'Class'                   Enum   0% 100%   0%     0 /  0%    11 /  1%   0% %%%%% Relabeled values in attribute 'Speaker Number'%    From: 0                       To: Andrew              %    From: 1                       To: Bill                %    From: 2                       To: David               %    From: 3                       To: Mark                %    From: 4                       To: Jo                  %    From: 5                       To: Kate                %    From: 6                       To: Penny               %    From: 7                       To: Rose                %    From: 8                       To: Mike                %    From: 9                       To: Nick                %    From: 10                      To: Rich                %    From: 11                      To: Tim                 %    From: 12                      To: Sarah               %    From: 13                      To: Sue                 %    From: 14                      To: Wendy               %%% Relabeled values in attribute 'Sex'%    From: 0                       To: Male