Self-similar group actions on ultragraphs and associated C ∗ C^{*} -algebras
Hossein Larki and Najmeh Rajabzadeh-Hasiri · 2025
Read Self-similar group actions on ultragraphs and associated C ∗ C^{*} -algebras by Hossein Larki and Najmeh Rajabzadeh-Hasiri (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-11949.txt.
Introduction
Hossein Larki and Najmeh Rajabzadeh-Hasiri extend a symmetry idea from graphs to a broader class called ultragraphs. Their 2025 paper defines what it means for a group to act on an ultragraph in a self-similar way, generalizing a notion due to Exel and Pardo. To each such action they attach a C*-algebra, an object blending algebra with analysis. They study those algebras through two models, one built from an inverse semigroup and one from a tight groupoid.
Catalog record
| CREATOR | Hossein Larki and Najmeh Rajabzadeh-Hasiri |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 100 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-11949
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Self-similar group actions on ultragraphs and associated C ∗ C^{*} -algebras Hossein Larki and Najmeh Rajabzadeh-Hasiri Abstract . As a generalization of the Exel-Pardo’s notion of self-similar graph, we introduce self-similar group actions on ultragraphs and their C ∗ C^{*} -algebras. We then approach to the C ∗ C^{*} -algebras by inverse semigroup and tight groupoid models. 2010 Mathematics Subject Classification 46L05 1. Introduction Ultragraph C^{*} -algebras were originally introduced by Tomforde [ 28 ] to give a unified framework for the graph C^{*} -algebras [ 17 , 18 ] and Exel-Laca algebras [ 9 ] . An ultragraph is in fact a generalized (directed) graph in which the source of each edge is a nonempty set of vertices rather than a single vertex. So, we expect that many ideas and results for the graph C^{*} -algebras in the literature could be extended (with more effort) to the ultragraph setting [ 21 , 25 , 28 , 29 ] . Furthermore, the algebraic analogue of the ultragraph C^{*} -algebras, namely ultragraph Leavitt path algebras, has attracted attension in recent years (see [ 2 , 4 , 13 , 14 ] among others). The notion of self-similar graph C^{*} -algebras was introduced by Exel and Pardo [ 10 ] to bring two important classes of C^{*} -algebras, the Katsura’s [ 15 ] and Nekrashevych’s algebras [ 26 ] under one theory like graph C^{*} -algebras. In particular, it contains all UCT Kirchberg C^{*} -algebras [ 10 ] and gives examples of important groupoid C^{*} -algebras with non-Hausdorff underlying groupoid [ 5 ] . The aim of this article is to generalize the Exel-Pardo C^{*} -algebras [ 10 ] to self-similar ultragraphs, and then give an inverse semigroup model for them. Note that self-similar graphs (G,E,\varphi) in [ 10 ] are assumed to satisfy \varphi(g,e)\cdot s(e)=g\cdot s(e) for all g\in G and e\in E^{1} . This condition is necessary to define a cyclic action of the group G on the path space E^{*} of E by g\cdot(\alpha\beta)=(g\cdot\alpha)(\varphi(g,\alpha)\cdot\beta) where \alpha,\beta\in E^{*} with s(\alpha)=r(\beta) . However, for the self-similar ultragraphs we just need (1.1) \varphi(g,e)\cdot s(e)\subseteq g\cdot s(e), because the source of each edge is a set of vertices (see Remark 3.3 below). In particular, by this weaker condition, we can easily construct finite self-similar ultragraphs which are not presented by self-similar graphs of [ 10 ] . Furthermore, we believe that our construction of self-similar ultragraph C^{*} -algebras can be also modified and extended to the labelled graph setting. (A labelled graph is another generalization of directed graphs in which both source and range of each edge are set of vertices.) This article is organized as follows. After reviewing necessary background in Section 2 , we introduce a self-similar ultragraph system (G,\mathcal{U},\varphi) in Section 3 , which is a self-similar group action G\curvearrowright\mathcal{U} on an ultragraph \mathcal{U}=(U^{0},\mathcal{U}^{1},r,s) equipped with a 1 -cocycle G\times\mathcal{U}^{1}\longrightarrow G satisfying condition ( 1.1 ) above, and we then associate to (G,\mathcal{U},\varphi) a universal C^{*} -algebra {\mathcal{O}}_{G,\mathcal{U}} satisfying certain conditions. In Section 4 , we consider some clases of examples. First, we show that for source-finite self-similar ultragraphs with a condition, the C^{*} -algebras {\mathcal{O}}_{G,\mathcal{U}} are included in the Exel-Pardo C^{*} -algebras of self-similar graphs. However, general C^{*} -algebras {\mathcal{O}}_{G,\mathcal{U}} are not; we construct a simple finite self-similar ultragraph which is not of the form self-similar graphs in [ 10 ] . Next, we define a self-similar analogue of Exel-Laca algebras [ 9 ] as another class of examples. In Section 5 , for each self-similar ultragraph (G,\mathcal{U},\varphi) we define an inverse semigroup {\mathcal{S}}_{G,\mathcal{U}} , and the rest of article is devoted to explain and analyze this inverse semigroup. Since ultragraphs contain generalized sets of vertices instead of singletons, elements of {\mathcal{S}}_{G,\mathcal{U}} are of the form (\alpha,A,g,\beta) , while we have triples (\alpha,g,\beta) in {\mathcal{S}}_{G,E} of [ 10 ] . It turns out in particular that {\mathcal{S}}_{G,\mathcal{U}} has a more complicated idempotent semilattice \mathcal{E}({\mathcal{S}}_{G,\mathcal{U}}) , among others. So, after recalling the description of the tight filter space {\widehat{\mathcal{E}}}_{\mathrm{tight}}({\mathcal{S}}_{G,\mathcal{U}}) from [ 4 ] and the construction of the tight groupoid {\mathcal{G}_{\mathrm{tight}}}({\mathcal{S}_{G,\mathcal{U}}}) in Section 6 , we compute in Section 7 the action {\mathcal{S}}_{G,\mathcal{U}}\curvearrowright{\widehat{\mathcal{E}}}_{\mathrm{tight}}({\mathcal{S}}_{G,\mathcal{U}}) carefully, which may be of interest even for ordinary ultragraphs \mathcal{U} . Then in Section 8 , we prove that the canonical representation \pi:{\mathcal{S}}_{G,\mathcal{U}}\longrightarrow{\mathcal{O}}_{G,\mathcal{U}} is a universal tight one, and therefore {\mathcal{O}}_{G,\mathcal{U}}\cong C^{*}_{\mathrm{tight}}({\mathcal{S}}_{G,\mathcal{U}})\cong C^{*}({\mathcal{G}_{\mathrm{tight}}}({\mathcal{S}_{G,\mathcal{U}}})) . Finally, in the last two sections, we characterize the Hausdorffness of the groupoid {\mathcal{G}_{\mathrm{tight}}}({\mathcal{S}_{G,\mathcal{U}}}) and the E^{*} -unitary property of {\mathcal{S}}_{G,\mathcal{U}} which generalize know results of [ 10 ] . We should notice that structural properties such as minimality and effectiveness of {\mathcal{G}_{\mathrm{tight}}}({\mathcal{S}_{G,\mathcal{U}}}) are more complicated than those in the self-similar graph setting, and we leave them to future works. 2. Preliminaries In this section we recall the necessary background for the rest of the paper. 2.1. Ultragraphs An ultragraph is a quadruple \mathcal{U}=(U^{0},\mathcal{U}^{1},r,s) , where U^{0} denoted the set of vertices, \mathcal{U}^{1} is the set of edges and r,s:\mathcal{U}^{1}\rightarrow P({U}^{0})\setminus\{\varnothing\} , describing the range and source maps, where P({U}^{0}) is the power set of {U}^{0} and r(e) is a singleton for every e\in{\mathcal{U}^{1}} . A source in \mathcal{U} is a vertex v\in{U}^{0} satisfying \ r^{-1}(\{v\})=\varnothing . Also, we say \mathcal{U} is row-finite whenever each vertex receives at most finitely many edges, that is \ r^{-1}(\{v\}) is a finite set for every v\in{U^{0}} . Remark 2.1 . We prefer to reverse the edge’s direction of ultragraphs in the literature [ 4 , 21 , 28 ] , which will be compatible with self-similar graphs and k -graphs in [ 10 , 11 , 19 , 23 , 24 , 20 ] . Standing assumption. Throughout the paper, we will assume that all ultragraphs \mathcal{U} are row-finite and without any source, meaning that the set r^{-1}(\{v\})=\{e\in\mathcal{U}^{1}\;:\;r(e)=\{v\}\} is nonempty and finite for all v\in U^{0} . Let \mathcal{U} be an ultragraph. We denote by {\mathcal{U}}^{0} the smallest subset of P({U}^{0}) containing \varnothing , \{v\} for all v\in{U}^{0} , s(e) for all e\in{\mathcal{U}}^{1} , and being closed under the finite set operations \cap , \cup and \backslash . A finite path in {\mathcal{U}} is either an element of {\mathcal{U}}^{0} or a sequence of edges \alpha=e_{1}\cdots e_{n} in {\mathcal{U}}^{1} such that r(e_{i+1})\subseteq s(e_{i}) for 1\leq i\leq n . Then the range of \alpha is defined by r(\alpha)=r(e_{1}) , the source of \alpha by s(\alpha)=s(e_{n}) , and the length of \alpha by |\alpha|=n . Each element {A}\in{\mathcal{U}}^{0} will be usually considered as a path of length zero in which case we set r(A)=s(A)=A . The set of all finite paths in \mathcal{U} is denoted by {\mathcal{U}}^{*}=\cup_{n=o}^{\infty}{\mathcal{U}}^{n} . An infinite path in \mathcal{U} is an infinite sequence of edges \alpha=e_{1}e_{2}\cdots , where r(e_{i+1})\subseteq s(e_{i}) for all i\geq 1 . The set of infinite paths in \mathcal{U} is denoted by {\mathcal{U}}^{\infty} . If \alpha and \beta are two paths in \mathcal{U}^{\geq 1} such that r(\beta)\subseteq s(\alpha) , we will denote by \alpha\beta the path obtained by juxtaposing \alpha and \beta . Remark 2.2 . In [ 28 ] , Tomforde defined {\mathcal{U}}^{0} to be the smallest subset of P({\mathcal{U}}^{0}) containing \{\{v\},\;s(e)\;:\;v\in U^{0}\;\text{and}\;e\in{\mathcal{U}}^{1}\} , which is only closed under the operation \cap and \cup . As noted in [ 16