Modeling and mining term association for improving biomedical information retrieval performance

BMC Bioinformatics

Table 8 Topic 200: ranking term associations

Rank	Term association	Term count	Communalities
1	high lupus serum	33	69.4
2	lupus protein serum	47	62.7
3	activ lupus serum	7	61.2
4	activ serum	118	60.0
5	activ associ diseas	124	59.9
6	associ diseas lupus	162	59.5
7	activ associ high	7	59.2
8	lupus serum	116	58.5
9	diseas high lupus	90	58.0
10	associ protein	20	58.0

Term associations are presented by applying the factor analysis based model: (1) the top 10 term associations are shown among $(C_{9}^{1} + C_{9}^{2} + C_{9}^{3})$ term associations generated by the proposed approach; (2) term count are computed from the top 1000 documents of the baseline; (3) the communality of each term association is calculated by the factor analysis based model; (4) the rank of term associations is given by their communalities.

ISSN: 1471-2105