--- title: Network Homology Hk-core Decomposition url: https://www.emergentmind.com/topics/network-homology-hk-core-decomposition type: topic --- # Network Homology Hk-core Decomposition Searching arXiv for the cited papers and closely related work on network homology, Hk-core decomposition, higher-order Laplacians, and generalized core decompositions. Network Homology Hk-core Decomposition denotes a family of decompositions that combine core-like pruning with homological descriptors of higher-order network structure. In one explicit formulation, the original simplicial network is the H0-core, and the H1-, H2-, H3-, and higher cores are obtained by iteratively deleting nodes and edges according to Betti numbers of node-neighbor subnetworks and a local characteristic number [2507.04948]. In a complementary spectral formulation, a network is modeled as a simplicial or cubical complex, the weighted \(k\)-Hodge Laplacian \(\mathcal L_k\) is formed, and the null space \(\ker \mathcal L_k \cong \mathcal H_k\) yields a homology embedding whose factorization isolates basic topological components and generators [2107.10970]. This suggests that the topic encompasses both a local homology-based pruning scheme and a global harmonic-subspace decomposition. ## 1. Mathematical setting and terminology The underlying network is typically represented as a simplicial complex \(K\) built from 0-simplices, 1-simplices, 2-simplices, 3-simplices, and higher simplices. If \(m_k\) denotes the number of \(k\)-simplices, then the Euler characteristic is \[ \chi = m_0 - m_1 + m_2 - m_3 + \cdots . \] For each \(k\), the chain space \(C_k\) is generated by the \(k\)-simplices, the boundary operator \(\partial_k:C_k\to C_{k-1}\) is represented by a boundary matrix \(B_k\), and if \(r_k=\operatorname{rank}(B_k)\), then \[ \beta_k = m_k - r_k - r_{k+1}. \] The Betti numbers \(\beta_k\) are the dimensions of the homology groups \[ H_k = \ker(\partial_k)/\operatorname{im}(\partial_{k+1}), \] with \(\beta_0\) counting connected components, \(\beta_1\) counting independent 1-dimensional cycles, and \(\beta_2\) counting 2-dimensional cavities [2507.04948]. A parallel formalism appears in the discrete Hodge-theoretic treatment of networks and manifolds. There, for simplicial or cubical complexes, one works with cochain spaces \(\mathcal C_k\), boundary matrices \(B_k\), and the \(k\)-th homology vector space \[ \mathcal H_k := \ker B_k / \mathrm{im}\, B_{k+1}, \] which is isomorphic to the null space of the \(k\)-Laplacian \(\mathcal L_k\) [2107.10970]. The terminology is not fully uniform across the literature. The expression “network homology Hk-core decomposition” is used explicitly for node-neighbor-subnetwork pruning [2507.04948], whereas the \(k\)-Laplacian work provides “essentially all the ingredients” to define and reason about a “network homology \(H_k\)-core decomposition” but does not itself use that term [2107.10970]. A plausible implication is that Hk-core decomposition is best understood as a homology-informed generalization of core decomposition rather than a single universally fixed construction. ## 2. Spectral homology embeddings and subspace factorization In the Hodge-theoretic formulation, the weighted normalized \(k\)-Laplacian is \[ \mathcal L_k = B_k^\top D_k^{-1} B_k + D_{k+1}^{-1} B_{k+1} B_{k+1}^\top = \mathcal B_k^\top \mathcal B_k + \mathcal B_{k+1}\mathcal B_{k+1}^\top, \] where \[ \mathcal B_\ell = W_{\ell-1}^{-1/2} B_\ell W_\ell^{1/2}. \] Its null space is the \(k\)-th homology vector space, and if \[ Y=[y_1,\dots,y_{\beta_k}] \in \mathbb R^{n_k\times \beta_k} \] is an orthonormal basis of \(\ker \mathcal L_k\), then the \(k\)-th homology embedding is defined by \[ \phi_k(\sigma_k)=Y_{\sigma_k,:}=\big(y_1(\sigma_k),\dots,y_{\beta_k}(\sigma_k)\big)^\top \in \mathbb R^{\beta_k}. \] Because the null-space basis is defined only up to an orthogonal transform, the embedding is defined only up to a global unitary transform in \(\mathbb R^{\beta_k}\) [2107.10970]. This construction generalizes the standard graph Laplacian embedding used in spectral clustering. For \(k=0\), \(\mathcal H_0\) encodes connected components; for \(k=1\), \(\mathcal H_1\) encodes loops; for \(k=2\), cavities; and in general each harmonic \(k\)-cochain corresponds to a homology class. The supports of these cochains identify representative cocycles for the corresponding holes [2107.10970]. The central structural result concerns connected sums. If \[ M = M_1 \sharp \cdots \sharp M_\kappa, \] then under assumptions that the complexes approximate the manifolds with isomorphic homology, that connected sum does not create or destroy \(k\)-homology, and that only a small fraction of \((k\pm1)\)-simplices are created or destroyed in gluing, the homology decomposes as \[ \mathcal H_k(M) \cong \bigoplus_{i=1}^{\kappa} \mathcal H_k(M_i), \] and the corresponding embedding behaves as a perturbation of the block-diagonal disjoint-union embedding. The resulting geometry is a union of linear subspaces, each associated with a distinct topological component. The paper therefore describes a “subspace clustering structure in the homology embedding” [2107.10970]. To extract these subspaces algorithmically, the homology basis \(Y\) is factorized by blind source separation. The proposed “Subspace identification” procedure constructs the complex, computes \(B_k\) and \(B_{k+1}\), forms \(\mathcal L_k\), computes the null-space basis \(Y\), and then applies ICA without sphering to obtain \[ Z = YW,\qquad W\in O(\beta_k). \] The columns of \(Z\) are treated as an independent homology basis, and each \(z_i\) can be interpreted as a “core” \(k\)-homology mode or basic generator [2107.10970]. This does not define Hk-cores in the node-peeling sense, but it furnishes a spectral notion of decomposition into topological cores. ## 3. Node-neighbor subnetworks and the H0/H1/H2/H3 hierarchy The node-neighbor formulation begins with a local subnetwork for each node \(i\), obtained from its neighbors and the simplices supported on that neighbor set. For each such node-neighbor subnetwork, the Betti numbers are computed exactly as for the full simplicial network, yielding a local tuple \[ (B_0,B_1,\dots,B_k)_i. \] The paper also introduces a local characteristic number. If the simplex counts of the deleted local piece associated with node \(i\) are \(m_k^{(i)}\), then \[ \chi_i^* = m_0^{(i)} - m_1^{(i)} + m_2^{(i)} - \cdots,\qquad \xi_i = 1 - \chi_i^*. \] The triple \[ \{n_i,(B_k)_i,\xi_i\} \] is the new node index, where \(n_i\) is the number of neighbors of node \(i\) in the original graph [2507.04948]. The decomposition itself is recursive. The H0-core is the original network. The H1-core is obtained by deleting nodes whose neighbor subnetworks have Betti numbers \[ (B_0=1,B_1=0,\dots,B_k=0)_i. \] A key result states that if a node-neighbor subnetwork has Betti numbers \((B_0,0,\dots,0)_i\), then deleting node \(i\) preserves the topology of the original network when \(B_0=1\), whereas if \(B_0>1\), deleting node \(i\) changes the Betti numbers and decreases the global \(B_1\) by \(B_0-1\) [2507.04948]. The H2-core is then obtained from the H1-core by deleting nodes with \[ (B_0>1,B_1=0,B_2=0,\dots,B_k=0)_i \] and, if \(B_1\) is still nonzero, deleting edges in subnetworks with \[ (B_0>1,B_1>0,B_2=0,\dots)_i \] that are attached to branches without 1-cycles. The H3-core is obtained from the H2-core by deleting nodes with \[ (B_0\le B_1, B_2=0,\dots,B_k=0)_i, \] and then deleting edges that connect to \(B_2\)-branches in a way that reduces \(B_2\) while retaining the higher-order structure. The construction continues inductively: the Hk-core is intended to retain the subnetwork that still carries homology in dimensions \(\ge k\) after lower-dimensional contributions have been removed [2507.04948]. Two local propositions explain why these deletions are topology-aware. First, if a neighbor subnetwork has a central node, then its Betti numbers are \((1,0,0,\dots,0)_i\). Second, deleting edges connected to nodes whose neighbor subnetworks have branches without \(B_k\) reduces \(B_k\). The method therefore peels off locally contractible pieces, then lower-order branches, then branches supporting progressively higher-dimensional holes [2507.04948]. This H0/H1/H2/H3 hierarchy is not degree-based. It is driven by the local homology of node-neighbor subnetworks and by repeated updates of the node indices after every node or edge deletion. Throughout the process, “the index of node involved in deleting edge needs to be updated in every step,” and the method is described as easy to implement in parallel [2507.04948]. ## 4. Relation to classical and generalized core decompositions The classical \(k\)-core of a graph is the maximal induced subgraph in which every vertex has degree at least \(k\). It is obtained by iteratively pruning vertices of degree \(0)_i\) changes the global \(B_3\) from 2 to 1. The paper then adopts a deletion-and-update strategy: after deleting 17 nodes with central-point neighborhoods and 33 nodes with \((1,0,0,0)_i\), the resulting H3-core has 15 nodes, \[ 6,10,12,15,16,20,22,23,24,27,41,53,54,59,60, \] with \[ m_0=15,\ m_1=67,\ m_2=129,\ m_3=116,\ m_4=49,\ m_5=12,\ m_6=1 \] and \[ (B_0,B_1,B_2,B_3)=(1,0,0,2). \] This preserves the two independent 3-dimensional cavities of the original network [2507.04948]. The spectral line of work supports a different set of applications. The homology embedding derived from \(\ker \mathcal L_k\) has been applied to the shortest homologous loop detection problem, which is NP-hard in general, and the spectral loop detection algorithm is reported to scale better than existing methods and to be effective on diverse data such as point clouds and images [2107.10970]. This suggests a practical division of labor: node-neighbor Hk-core decomposition localizes higher-order cavities, whereas null-space factorization yields harmonic generators and homology modes. ## 6. Interpretation, limitations, and open issues A common misconception is to treat Hk-core decomposition as a direct synonym for graph \(k\)-core. The literature does not support that identification. Classical \(k\)-core is degree-threshold peeling on the 1-skeleton; Hk-core decomposition uses local Betti numbers, Euler-characteristic-derived node indices, and node-neighbor simplicial subnetworks [2507.04948]. Another misconception is to equate Hk-core decomposition with persistent homology. The node-neighbor method operates at a single combinatorial scale and is presented as a way to simplify homology calculation; it is not a substitute for persistence but a complementary tool to structurally localize and simplify cavity search [2507.04948]. The spectral interpretation also has explicit assumptions. Its perturbation theorem requires that the simplicial or cubical complex approximate the underlying manifold with isomorphic homology, that connected sum not create or destroy \(k\)-homology, and that only a small fraction of \((k\pm1)\)-cells be created or destroyed in gluing. The ICA-based identification of independent harmonic modes is data-driven and heuristic; the paper states that formal guarantees that each component aligns exactly with a single homology generator are not fully proven, though the behavior is supported empirically and by perturbation analysis [2107.10970]. The node-neighbor formulation has its own sensitivities. The cat cortical example shows that a naive retention strategy can change the target Betti number, whereas deletion with repeated local updates preserves the intended homology [2507.04948]. This indicates that the exact deletion rules and update order are structural components of the method rather than incidental implementation details. A plausible implication is that Hk-core decomposition is best viewed as an algorithmic homology-preserving reduction scheme, not merely as a static invariant. The broader research landscape suggests several convergences. Generalized cores indexed by distance, clique support, or hyperedge size already provide multi-scale and higher-order decompositions [1904.07262; 1704.00386; 2301.06712]. Hodge-theoretic embeddings furnish harmonic bases for \(\mathcal H_k\) [2107.10970]. Hk-core decomposition occupies the intersection of these lines by using topological rather than purely combinatorial support to define the “core” of a network. In that sense, it reframes core decomposition from a degree-centric concept into a homology-centric one. Source: https://www.emergentmind.com/topics/network-homology-hk-core-decomposition