Kishida, K.: High-speed rough clustering for very large document collections (2010)
0.01
0.006830811 = product of:
0.034154054 = sum of:
0.034154054 = weight(_text_:it in 3463) [ClassicSimilarity], result of:
0.034154054 = score(doc=3463,freq=4.0), product of:
0.15115225 = queryWeight, product of:
2.892262 = idf(docFreq=6664, maxDocs=44218)
0.052260913 = queryNorm
0.22595796 = fieldWeight in 3463, product of:
2.0 = tf(freq=4.0), with freq of:
4.0 = termFreq=4.0
2.892262 = idf(docFreq=6664, maxDocs=44218)
0.0390625 = fieldNorm(doc=3463)
0.2 = coord(1/5)
- Abstract
- Document clustering is an important tool, but it is not yet widely used in practice probably because of its high computational complexity. This article explores techniques of high-speed rough clustering of documents, assuming that it is sometimes necessary to obtain a clustering result in a shorter time, although the result is just an approximate outline of document clusters. A promising approach for such clustering is to reduce the number of documents to be checked for generating cluster vectors in the leader-follower clustering algorithm. Based on this idea, the present article proposes a modified Crouch algorithm and incomplete single-pass leader-follower algorithm. Also, a two-stage grouping technique, in which the first stage attempts to decrease the number of documents to be processed in the second stage by applying a quick merging technique, is developed. An experiment using a part of the Reuters corpus RCV1 showed empirically that both the modified Crouch and the incomplete single-pass leader-follower algorithms achieve clustering results more efficiently than the original methods, and also improved the effectiveness of clustering results. On the other hand, the two-stage grouping technique did not reduce the processing time in this experiment.