scientific article; zbMATH DE number 1329697

From MaRDI portal
Publication:4258052





For \(G\) a locally compact group with (left) Haar measure the commutant of the von Neumann algebra generated by the right regular representation can be identified with the von Neumann algebra generated by the left regular representation (the commutant theorem). The author states something like that when there are topological groups \(G\) and \({\widetilde{G}}\) such that \({\widetilde{G}}\) is the completion of \(G\) with respect to a suitable metric; for any measure on \({\widetilde{G}}\) which is G-quasi-invariant with respect to the right action of \({\widetilde{G}}\) (on itself) one can construct an analogue of the right regular representation in the usual way and when the measure is also quasi-invariant with respect to the inverse mapping one can construct the corresponding left regular representation. He states without proof that the commutant theorem then holds. The author's main theorem concerns \({\widetilde{G}}\) as the group of upper triangular infinite matrices and \(G\) the subgroup with finite entries. The Gaussian measure chosen is the projection to \({\widetilde{G}}\) of the Gaussian measure on the group of all infinite matrices with positive entries, as defined by the author in [Funct. Anal. Appl. 24, 243-245 (1990); translation from Funkts. Anal. Prilozh. 24, 82-83 (1990; Zbl 0780.22007)]. Analogously to the use by \textit{S. Kakutani} [Ann. Math. (2) 49, 214-224 (1948; Zbl 0030.02303)] of Hellinger's integral, the author (loc. cit.) showed that if \(\sum_{m=n+1}^{\infty} { {b_{km}} \over {b_{nm}}} < \infty\) for all \(k \in {\mathbb N}\) ensures quasi-invariance with respect to the inverse mapping and the construction of left and right regular representations. He proves here that the commutant theorem holds. Besides the groups of infinite matrices the author refers to current groups, i.e., where the groups \(G\) and \({\widetilde{G}}\) consist of smooth sections \(X \to U\) for \(U\) a Lie group. His use of the term ``infinite dimensional group is sloppy. For the entire collection see [Zbl 0914.00060].











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4258052)