OpenCog ULL - clustering	2017-12-23
Vector space	Gutenberg_16_10
Algorithm	KMeans - sklearn.cluster
Number of clusters	20
Lexicon file	Gut_16_10_lexicon.csv
Lexicon size (words)	6855
AVERAGE METRICS FOR ALL CLUSTERS	-
Average top cluster word similarity to centroid	0.42
Average top 10 words similarity to centroid	0.392
Average cluster words similarity to centroid	0.251
Similarity weighted by words frequency within cluster (SWFC)	0.228
Similarity weighted by words frequency within lexicon (SWFL)	0.004
METRICS FOR CLUSTERS	-
Cluster sizes (words)	380, 314, 1309, 307, 322, 251, 341, 216, 296, 294, 307, 239, 250, 321, 373, 281, 225, 324, 221, 284
Cluster top words	duel, disembodied, emma, neighbours, triumphant, select, distributed, knitting, strove, prof, mesmerism, blurted, fates, youngest, prevented, citizenship, destruction, haughtily, weariness, shoemaker
Cluster top word similarities to centroids	0.453, 0.465, 0.115, 0.476, 0.432, 0.423, 0.447, 0.422, 0.417, 0.442, 0.414, 0.437, 0.499, 0.393, 0.373, 0.45, 0.439, 0.43, 0.445, 0.433
Cluster top 10 words avg similarities to centroids	0.433, 0.431, 0.109, 0.431, 0.407, 0.388, 0.42, 0.373, 0.405, 0.386, 0.395, 0.371, 0.47, 0.375, 0.344, 0.433, 0.42, 0.427, 0.424, 0.399
Cluster words avg similarities to centroids	0.258, 0.272, 0.052, 0.285, 0.26, 0.274, 0.271, 0.246, 0.225, 0.249, 0.264, 0.245, 0.303, 0.235, 0.217, 0.3, 0.293, 0.234, 0.296, 0.249
Similarities weighted by words frequency within cluster (SWFC)	0.246, 0.246, 0.035, 0.253, 0.232, 0.237, 0.246, 0.226, 0.213, 0.228, 0.235, 0.214, 0.272, 0.219, 0.217, 0.264, 0.249, 0.226, 0.264, 0.236
Similarities weighted by words frequency within lexicon (SWFL)	0.002, 0.003, 0.027, 0.003, 0.003, 0.003, 0.002, 0.002, 0.002, 0.002, 0.003, 0.003, 0.002, 0.002, 0.009, 0.003, 0.004, 0.001, 0.003, 0.002
SUPPLEMENTAL FILES	CONTENT
Gut_16_10_20_cluster_top_words.txt	Cluster top 10 words statistics
Gut_16_10_20_cluster_words.txt	Cluster words, sorted by similarities to cluster centroids
Gut_16_10_20_cluster_similarities.txt	Cluster word similarities to cluster centroids