Ossorio, P., \Classication space: a multivariate procedure for automatic document indexing and retrieval", Multivariate Behavior Research, 1966, pp. 479-524.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... have been issued a patent for the basic algorithm, however Bellcore (their employer at the time of ling) is the owner [Deerwester et al. 1988] 5 the need for tedious data gathering techniques, requiring collection of thousands of similarity judgements from humans [Borko, Bernick 1963] [Ossorio 1966]. According to the inventors of LSI, the three strengths of the algorithm are: it involves a high dimensional representation, which allows one to better represent a wide range of semantic relations; both terms and text objects are explicitly represented in the same space; and objects can ....

Ossorio, P., \Classication space: a multivariate procedure for automatic document indexing and retrieval", Multivariate Behavior Research, 1966, pp. 479-524.

.... serious problems associated with factor analytic approaches are: 1) high computational expense, 2) diculty impossibility of implementation of all but low dimensional representations, and (3) the need to gather collections of thousands of similarity judgements from humans [Borko, Bernick 1963] [Ossorio 1966]. LSI does not su er from these drawbacks, and has three additional strengths, namely: 1) it involves a high dimensional representation, which allows one to better represent a wide range of semantic relations; 2) both terms and text objects are explicitly represented in the same space; and (3) ....

Ossorio, P., \Classication space: a multivariate procedure for automatic document

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC