alphaFeedback

WorksMathematics

Upper kk-star-forming sets and kk-independence in bipartite graphs

AI contributions
  • Anthropic Claude Code Opus 5.5 Ultracode
View details
Ancillary data
2 code links
View

Abstract

Sahbi conjectured that the kk-independence number βk\beta_k and the upper kk-star-forming number \SF_k agree on bipartite graphs, and in particular on chain graphs, and asked whether \beta_3=\SF_3 on triangle-free graphs. For every k≥3k\ge3 we give a connected bipartite graph on 3k+33k+3 vertices with βk=2k+2\beta_k=2k+2 and \SF_k\ge3k. On chain graphs, however, \beta_k=\SF_k for every kk. We also show, answering another of his questions, that deciding whether a bipartite graph has a minimal 22-star-forming set with at least tt vertices is NP-complete.

Introduction

Let GG be a finite simple graph and kk a positive integer. A set II of vertices is kk-independent if the induced subgraph G[I]G[I] has maximum degree less than kk, and βk(G)\beta_k(G) is the largest size of such a set; this notion goes back to Fink and Jacobson [7], see the survey [5] and [3]. Chellali and Favaron [4] called a set SS kk-star-forming if for every vertex v∉Sv \notin S the graph G[S∪{v}]G[S \cup\{v\}] contains a copy of the star K1,kK_{1,k} that contains vv, and they showed that βk(G)≤SFk(G)\beta_k(G) \le\mathrm{SF}_k(G), where SFk(G)\mathrm{SF}_k(G) is the largest size of a minimal kk-star-forming set.1 For k=1k = 1

Email address: [email protected] (Deep Bhattacharjee) these are the independence number α\alpha and the upper domination number Γ\Gamma, and the theorem of Cockayne, Favaron, Payan and Thomason [6] that Γ=α\Gamma= \alpha for bipartite graphs is the first case of the following conjecture of Sahbi.2

Question 1.1 ([13 Conjecture 4.1]). Is βk(G)=SFk(G)\beta_k(G) = \mathrm{SF}_k(G) for every bipartite graph GG and every k≥1k \ge1?

Sahbi proved the equality for complete bipartite graphs and every kk, and for chain graphs when k=2k = 2; he gave a triangle-free graph on ten vertices with β4<SF4\beta_4 < \mathrm{SF}_4, and asked [13 Problem 7.3] whether β3=SF3\beta_3 = \mathrm{SF}_3 for every triangle-free graph. The case k=2k = 2 of the conjecture was refuted by a ten-vertex bipartite graph with β2=5\beta_2 = 5 and SF2=6\mathrm{SF}_2 = 6, posted on 9 October 2026 to a public repository of refutations [1].3 We show that the conjecture fails for every larger kk as well.

Theorem 1.2. For every k≥3k \ge3 the connected bipartite graph GkG_k of § 2, on 3k+33k + 3 vertices, satisfies

βk(Gk)=2k+2andSFk(Gk)≥3k.\beta_k(G_k) = 2k + 2 \qquad\text{and} \qquad\mathrm{SF}_k(G_k) \ge3k.

In particular SFk(Gk)−βk(Gk)≥k−2\mathrm{SF}_k(G_k) - \beta_k(G_k) \ge k - 2, and the answer to Problem 7.3 of [13] is negative already for bipartite graphs: G3G_3 has twelve vertices and β3(G3)=8<9≤SF3(G3)\beta_3(G_3) = 8 < 9 \le\mathrm{SF}_3(G_3). Together with the example for k=2k = 2, the equality βk=SFk\beta_k = \mathrm{SF}_k holds on all bipartite graphs only for k=1k = 1, which answers Sahbi’s Problem 9.4 for the class of bipartite graphs.

The equality does survive on the more rigid class of chain graphs, the bipartite graphs in which the neighbourhoods of the vertices on one side are nested. Sahbi proved β2=SF2\beta_2 = \mathrm{SF}_2 for chain graphs and conjectured [13 Conjecture 6.4] that βk=SFk\beta_k = \mathrm{SF}_k for chain graphs and all k≥3k \ge3; this is his Problem 9.1. We prove it.

Theorem 1.3. For every chain graph GG and every k≥2k \ge2, βk(G)=SFk(G)\beta_k(G) = \mathrm{SF}_k(G).

The proof in § 4 shows that in a chain graph a minimal kk-star-forming set has at most one vertex with kk or more neighbours inside it, and that this vertex can be exchanged for a vertex outside the set. The graphs GkG_k show that some such rigidity is needed: in GkG_k all k−1k-1 hubs have 2k−12k-1 neighbours in the set SS of Lemma 3.1.

Sahbi also asked [13 Problem 9.3] for the complexity of computing SF2\mathrm{SF}_2 on bipartite graphs, and in particular whether it is NP-complete to decide, for a bipartite graph GG and an integer tt, whether GG has a minimal 2-star-forming set with at least tt vertices. It is. For a graph HH let H♯H^\sharp be the graph obtained from HH by replacing every edge with a path of length four and attaching a pendant vertex to every vertex of HH (§ 5).

Theorem 1.4. Let HH be a graph with nn vertices, mm edges and independence number α(H)\alpha(H). Then H♯H^\sharp is a bipartite graph on 2n+3m2n+3m vertices, and

β2(H♯)=SF2(H♯)=n+2m+α(H).\beta_2(H^\sharp)=\mathrm{SF}_2(H^\sharp)=n+2m+\alpha(H).

Hence it is NP-complete to decide whether a bipartite graph has a minimal 2-star-forming set with at least tt vertices, and this remains so for planar bipartite graphs.

The graphs H♯H^\sharp satisfy Sahbi’s equality, so the hardness does not come from a gap between the two numbers. The same identity gives another proof that computing β2\beta_2, the dissociation number, is NP-hard on bipartite graphs, a result of Yannakakis [14].

The graphs GkG_k

We write NS(v)N_S(v) for the set of neighbours of vv in SS and dS(v)=∣NS(v)∣d_S(v)=\lvert N_S(v)\rvert. A copy of K1,kK_{1,k} in G[S∪{v}]G[S\cup\{v\}] containing v∉Sv\notin S has vv as its centre or as a leaf, which gives the following local form of the definition [13 Lemma 2.2].

Lemma 2.1. A set SS is kk-star-forming if and only if every vertex v∉Sv\notin S satisfies dS(v)≥kd_S(v)\ge k or has a neighbour u∈Su\in S with dS(u)≥k−1d_S(u)\ge k-1. Every superset of a kk-star-forming set is kk-star-forming, so a kk-star-forming set SS is minimal if and only if S∖{s}S\setminus\{s\} is not kk-star-forming for each s∈Ss\in S.

Fix k≥2k\ge2. The graph GkG_k has the 3k+33k+3 vertices

h1,…,hk−1,u1,…,uk−1,w1,…,wk−1,x0,p,c1,c2,q1,q2.h_1,\ldots,h_{k-1},\qquad u_1,\ldots,u_{k-1},\qquad w_1,\ldots,w_{k-1},\qquad x_0,p,c_1,c_2,q_1,q_2.

Write H={h1,…,hk−1}H=\{h_1,\ldots,h_{k-1}\} for the hubs, X1={u1,…,uk−1}X_1=\{u_1,\ldots,u_{k-1}\}, X2={w1,…,wk−1}X_2=\{w_1,\ldots,w_{k-1}\} and X=X1∪X2∪{x0}X=X_1\cup X_2\cup\{x_0\}. The edges are

  • hixh_i x for every hub hih_i and every x∈Xx \in X;

  • x0px_0p, c1q1c_1q_1 and c2q2c_2q_2;

  • c1ujc_1u_j and c2wjc_2w_j for 1≤j≤k−11 \leq j \leq k - 1.

So GkG_k has (k−1)(2k−1)+2(k−1)+3=2k2−k+2(k-1)(2k-1)+2(k-1)+3=2k^2-k+2 edges. It is connected, and it is bipartite with parts H∪{p,c1,c2}H \cup\{p,c_1,c_2\} and X∪{q1,q2}X \cup\{q_1,q_2\}. Figure 1 shows G3G_3, which has 12 vertices and 17 edges.

The graph $G_3$

Figure 1. The graph G3G_3. The nine shaded vertices form a minimal 3-star-forming set SS, while every 3-independent set has at most eight vertices. The thick edges are the three stars K1,3K_{1,3} that place pp, q1q_1 and q2q_2 in stars over SS.

The vertices x0x_0, c1c_1 and c2c_2 each see exactly k−1k-1 vertices of H∪XH \cup X: the hubs, the block X1X_1 and the block X2X_2, respectively. Each of them also carries a pendant vertex, and the three pendant vertices pp, q1q_1, q2q_2 are left outside the set SS below. Each pendant vertex is a leaf of a star whose centre and other leaves are forced into SS, and this is what makes SS minimal although SS is far from kk-independent: a hub has 2k−12k-1 neighbours in SS.

Proof of Theorem 1.2

Lemma 3.1. For k≥2k \geq2 the set S=V(Gk)∖{p,q1,q2}S=V(G_k)\setminus\{p,q_1,q_2\} is a minimal kk-star-forming set of GkG_k with ∣S∣=3k|S|=3k.

Proof. Clearly ∣S∣=3(k−1)+3=3k|S|=3(k-1)+3=3k. In SS we have NS(x0)=HN_S(x_0)=H, NS(c1)=X1N_S(c_1)=X_1 and NS(c2)=X2N_S(c_2)=X_2, so dS(x0)=dS(c1)=dS(c2)=k−1d_S(x_0)=d_S(c_1)=d_S(c_2)=k-1. The only vertices outside SS are pp, q1q_1, q2q_2, and their neighbours x0x_0, c1c_1, c2c_2 lie in SS; by Lemma 2.1, SS is kk-star-forming.

For minimality we use Lemma 2.1 again. The pendant vertices have degree 1<k1 < k, so after the deletion of a vertex ss from SS a pendant vertex is still in a star only if its unique neighbour remains in S∖{s}S \setminus\{s\} with at least k−1k-1 neighbours there. If ss is a hub, then x0x_0 keeps only k−2k-2 neighbours, so pp lies in no star. If s=x0s=x_0, then pp has no neighbour left. Likewise deleting a vertex of X1X_1, or c1c_1, leaves q1q_1 outside every star, and deleting a vertex of X2X_2, or c2c_2, does the same to q2q_2. These cases cover all of SS.

Lemma 3.2. For k≥2k \ge2, βk(Gk)=2k+2\beta_k(G_k)=2k+2.

Proof. The set I0=X∪{p,q1,q2}I_0=X \cup\{p,q_1,q_2\} induces the single edge x0px_0p together with isolated vertices, so it is kk-independent, and ∣I0∣=2(k−1)+1+3=2k+2|I_0|=2(k-1)+1+3=2k+2.

Conversely, let II be kk-independent and put t=∣I∩H∣t=|I \cap H| and s=∣I∩X∣s=|I \cap X|. If t=0t=0, then II lies in Gk−HG_k-H, which is the disjoint union of the edge x0px_0p and the two stars with centres c1,c2c_1,c_2 and leaf sets X1∪{q1}X_1 \cup\{q_1\}, X2∪{q2}X_2 \cup\{q_2\}, each a copy of K1,kK_{1,k}. A kk-independent set meets K1,kK_{1,k} in at most kk vertices, since if it contains the centre it must omit a leaf; hence ∣I∣≤2+k+k|I| \le2+k+k.

Suppose t≥1t \ge1. A hub in II is adjacent to every vertex of I∩XI \cap X, so s≤k−1s \le k-1. Besides HH and XX, at most the five vertices p,c1,c2,q1,q2p,c_1,c_2,q_1,q_2 remain, so ∣I∣≤t+s+5|I| \le t+s+5, which is at most 2k+22k+2 when t≤k−2t \le k-2. Finally let t=k−1t=k-1. Every x∈I∩Xx \in I \cap X is adjacent to all k−1k-1 hubs, all of which lie in II, so its remaining neighbour among p,c1,c2p,c_1,c_2 (namely pp for x0x_0, c1c_1 on X1X_1 and c2c_2 on X2X_2) is not in II. If s≥1s \ge1 this removes at least one of p,c1,c2p,c_1,c_2, and ∣I∣≤(k−1)+(k−1)+2+2=2k+2|I| \le(k-1)+(k-1)+2+2=2k+2. If s=0s=0, then ∣I∣≤(k−1)+5≤2k+2|I| \le(k-1)+5 \le2k+2.

Proof of Theorem 1.2. By Lemma 3.1, SFk(Gk)≥∣S∣=3k\mathrm{SF}_k(G_k) \ge|S|=3k, and by Lemma 3.2, βk(Gk)=2k+2\beta_k(G_k)=2k+2, which is smaller than 3k3k once k≥3k \ge3.

Chain graphs

Let GG be a chain graph with parts AA and BB, so that the neighbourhoods N(a)N(a), a∈Aa \in A, form a chain under inclusion. Then the neighbourhoods N(b)N(b), b∈Bb \in B, form a chain as well.4 Two vertices on the same side therefore satisfy N(u)⊆N(v)N(u) \subseteq N(v) or N(v)⊆N(u)N(v) \subseteq N(u), and N(u)⊆N(v)N(u) \subseteq N(v) implies dS(u)≤dS(v)d_S(u) \le d_S(v) for every set SS. We say that SS covers a vertex v∉Sv \notin S if vv satisfies the condition of Lemma 2.1. Deleting a vertex ss from SS changes dSd_S only on the neighbours of ss, which lie on the other side.

The exchange step works in every graph.

Lemma 4.1. Let SS be a minimal kk-star-forming set of a graph GG, k≥1k \ge1, such that at most one vertex v∈Sv \in S has dS(v)≥kd_S(v) \ge k. Then βk(G)≥∣S∣\beta_k(G) \ge|S|.

Proof. If no vertex of SS has dS(v)≥kd_S(v) \ge k, then SS itself is kk-independent. Otherwise let hh be the vertex with dS(h)≥kd_S(h) \ge k. By Lemma 2.1 some vertex z∉S∖{h}z \notin S \setminus\{h\} is not covered by S∖{h}S \setminus\{h\}: it has at most k−1k-1 neighbours in S∖{h}S \setminus\{h\}, and each of these has at most k−2k-2 neighbours in S∖{h}S \setminus\{h\}. As hh has dS(h)≥kd_S(h) \ge k neighbours in S∖{h}S \setminus\{h\}, we have z≠hz \ne h, so z∉Sz \notin S. In I=(S∖{h})∪{z}I = (S \setminus\{h\}) \cup\{z\} the vertex zz has at most k−1k-1 neighbours, each neighbour of zz has at most (k−2)+1(k-2)+1, and every other vertex vv has at most dS(v)≤k−1d_S(v) \le k-1, since v≠hv \ne h. Hence II is kk-independent, and ∣I∣=∣S∣|I| = |S|.

Lemma 4.2. Let GG be a chain graph, k≥2k \ge2, and SS a minimal kk-star-forming set. If h∈Sh \in S has dS(h)≥kd_S(h) \ge k, then dS(y)≤k−2d_S(y) \le k-2 for every y∈Sy \in S on the other side.

Proof. We first show that every neighbour yy of hh in SS has a neighbour w∉Sw \notin S such that dS(u)≤k−2d_S(u) \le k-2 for every neighbour u≠yu \ne y of ww in $S. Some vertex w∉S∖{y}w \notin S \setminus\{y\} is not covered by S∖{y}S \setminus\{y\}. It is not yy, since hh is a neighbour of yy with dS(h)−1≥k−1d_S(h)-1 \ge k-1 neighbours in S∖{y}S \setminus\{y\}; so w∉Sw \notin S, and SS covers ww. Suppose ww lies on the side of yy. Then ww is not adjacent to yy, so dS(w)=dS∖{y}(w)<kd_S(w) = d_{S \setminus\{y\}}(w) < k, and ww has a neighbour u∈Su \in S with dS(u)≥k−1d_S(u) \ge k-1, and uu lies on the side of hh. If N(u)⊆N(h)N(u) \subseteq N(h), then hh is a neighbour of ww with at least k−1k-1 neighbours in S∖{y}S \setminus\{y\}; if N(h)⊆N(u)N(h) \subseteq N(u), then dS(u)≥dS(h)≥kd_S(u) \ge d_S(h) \ge k and uu keeps at least k−1k-1 neighbours in S∖{y}S \setminus\{y\}. Either way S∖{y}S \setminus\{y\} covers ww, which is impossible. So ww lies on the side of hh, and its neighbours, lying on the side of yy, have the same degree in S∖{y}S \setminus\{y\} as in SS. If ww were not adjacent to yy, it would also keep its own degree, and S∖{y}S \setminus\{y\} would cover ww as SS does. Hence ww is adjacent to yy, and since S∖{y}S \setminus\{y\} does not cover ww, every neighbour u≠yu \ne y of ww in SS has dS(u)≤k−2d_S(u) \le k-2.

Now hh has two neighbours a≠ba \ne b in SS, and we may assume N(a)⊆N(b)N(a) \subseteq N(b). Let ww be the vertex just found for aa. Then ww is adjacent to bb, so dS(b)≤k−2d_S(b) \le k-2. Let y∈Sy \in S lie on the side of aa and bb. If N(y)⊆N(b)N(y) \subseteq N(b), then $d_S(y) \le d_S(b) \le k−2k-2. If N(b)⊆N(y)N(b) \subseteq N(y), then ww is adjacent to yy, so dS(y)≤k−2d_S(y) \le k-2 when y≠ay \ne a, while dS(a)≤dS(b)d_S(a) \le d_S(b). □\square

Lemma 4.3. Let GG be a chain graph, k≥2k \ge2, and SS a minimal kk-star-forming set. At most one vertex v∈Sv \in S has dS(v)≥kd_S(v) \ge k.

Proof. Suppose that h≠th \ne t are two such vertices. By Lemma 4.2 they lie on the same side, and we may assume N(t)⊆N(h)N(t) \subseteq N(h). Some vertex z∉S∖{t}z \notin S \setminus\{t\} is not covered by S∖{t}S \setminus\{t\}, and z≠tz \ne t since dS∖{t}(t)=dS(t)≥kd_{S\setminus\{t\}}(t) = d_S(t) \ge k. So z∉Sz \notin S, and SS covers zz. If zz lies on the side of tt, then zz is not adjacent to tt, so dS(z)=dS∖{t}(z)<kd_S(z) = d_{S\setminus\{t\}}(z) < k, and zz has a neighbour u∈Su \in S with dS(u)≥k−1d_S(u) \ge k-1; but uu lies on the other side from hh, and dS(u)≤k−2d_S(u) \le k-2 by Lemma 4.2. If zz lies on the other side, then zz is not adjacent to hh, which has dS(h)≥kd_S(h) \ge k neighbours in S∖{t}S \setminus\{t\}; hence zz is not adjacent to tt either, as N(t)⊆N(h)N(t) \subseteq N(h). Then neither zz nor any of its neighbours, which lie on the side of tt, changes its degree when tt is deleted, and S∖{t}S \setminus\{t\} covers zz as SS does. Both cases are impossible. □\square

Proof of Theorem 1.3. We have βk(G)≤SF⁡k(G)\beta_k(G) \le\operatorname{SF}_k(G) by [4]. Conversely, every minimal kk-star-forming set SS satisfies ∣S∣≤βk(G)|S| \le\beta_k(G) by Lemmas 4.3 and 4.1. □\square

For k=1k = 1 the equality holds on all bipartite graphs [6]; the argument above needs k≥2k \ge2, since in Lemma 4.2 the vertex hh must have two neighbours in SS.

Complexity

Let HH be a graph with vertex set VV, nn vertices and mm edges, and write each edge as an ordered pair e=(ue,ve)e = (u_e, v_e) of its ends, in either order. The graph H♯H^\sharp has the vertices of HH, a new vertex v′v' for every v∈Vv \in V and three new vertices ae,be,cea_e, b_e, c_e for every edge ee. Its edges are vv′vv' for v∈Vv \in V and the four edges of the path ueaebeceveu_ea_eb_ec_ev_e for every edge ee of HH. So H♯H^\sharp has 2n+3m2n + 3m vertices; it is bipartite with parts V∪{be}V \cup\{b_e\} and {v′}∪{ae,ce}\{v'\} \cup\{a_e, c_e\}, and it is planar when HH is. Figure 2 shows K3♯K^\sharp_3.

The graph $K_3^\sharp$

Figure 2. The graph K3♯K_3^\sharp, with the edges ordered as (x,y)(x,y), (y,z)(y,z), (x,z)(x,z). The ten shaded vertices form the set J({x})J(\{x\}) of Lemma 5.1. It is a largest 2-independent set and a largest minimal 2-star-forming set, and 10=n+2m+α(K3)10=n+2m+\alpha(K_3).

Lemma 5.1. For every independent set II of HH the set

J(I)=I∪{v′:v∈V}∪{be:e∈E(H)}∪{ae:ue∉I}∪{ce:ue∈I}J(I) = I \cup\{v' : v \in V\} \cup\{b_e : e \in E(H)\} \cup\{a_e : u_e \notin I\} \cup\{c_e : u_e \in I\}

is 22-independent in H♯H^\sharp and has n+2m+∣I∣n + 2m + |I| vertices.

Proof. The size is clear, since J(I)J(I) contains exactly one of aea_e and cec_e for each edge ee. We check that every vertex of J(I)J(I) has at most one neighbour in J(I)J(I). A vertex v′v' has only one neighbour, and beb_e has exactly one of its neighbours ae,cea_e,c_e in J(I)J(I). Let v∈Iv\in I. For an edge ee with ue=vu_e=v the set J(I)J(I) contains cec_e and not aea_e, and for an edge ee with ve=vv_e=v we have ue∉Iu_e\notin I, because II is independent, so J(I)J(I) contains aea_e and not cec_e; hence v′v' is the only neighbour of vv in J(I)J(I). If ae∈J(I)a_e\in J(I), then ue∉Iu_e\notin I, so beb_e is its only neighbour in J(I)J(I). If ce∈J(I)c_e\in J(I), then ue∈Iu_e\in I, so ve∉Iv_e\notin I, and again beb_e is its only neighbour in J(I)J(I).

For the upper bound we use Lemma 2.1 with k=2k=2: a set SS covers w∉Sw\notin S if dS(w)≥2d_S(w)\ge2 or ww has a neighbour x∈Sx\in S with dS(x)≥1d_S(x)\ge1. Let s∈Ss\in S. If a vertex w∉Sw\notin S is neither adjacent to ss nor at distance two from it, then dS∖{s}(w)=dS(w)d_{S\setminus\{s\}}(w)=d_S(w), and every neighbour xx of ww satisfies x≠sx\ne s and dS∖{s}(x)=dS(x)d_{S\setminus\{s\}}(x)=d_S(x), so S∖{s}S\setminus\{s\} covers ww whenever SS does. Hence, to see that S∖{s}S\setminus\{s\} is still 2-star-forming, we only have to check ss and the vertices outside SS within distance two of ss.

Lemma 5.2. Let SS be a minimal 2-star-forming set of H♯H^\sharp, and let YY be the set of vertices v∈Vv\in V with v∈Sv\in S and v′∈Sv'\in S.

(a) Every v∈Vv\in V satisfies v∈Sv\in S or $v'\in S.

(b) For no edge ee do all three of ae,be,cea_e,b_e,c_e lie in SS.

(c) If both ends of an edge ee lie in YY, then at most one of ae,be,cea_e,b_e,c_e lies in SS.

Proof. (a) The vertex v′v' has the single neighbour vv, so if v′∉Sv' \notin S then SS covers v′v' only through v∈Sv \in S.

(b) Suppose that ae,be,ce∈Sa_e,b_e,c_e \in S, and put S′=S∖{be}S' = S \setminus\{b_e\}. The vertex beb_e has its two neighbours ae,cea_e,c_e in S′S'. The vertices outside SS at distance two from beb_e are among ueu_e and vev_e. If ue∉Su_e \notin S, then ue′∈Su'_e \in S by (a), so ueu_e has the two neighbours ue′u'_e and aea_e in S′S', and likewise for vev_e. So S′S' is 2-star-forming, against the minimality of SS.

(c) Let ue,ve∈Yu_e,v_e \in Y and suppose that two of ae,be,cea_e,b_e,c_e lie in SS; by (b) the third does not. Since ue′∈Su'_e \in S and ve′∈Sv'_e \in S, the vertices ueu_e and vev_e keep at least one neighbour in SS after the deletion of any of ae,be,cea_e,b_e,c_e, so every vertex outside SS that is adjacent to ueu_e or vev_e stays covered. If be∉Sb_e \notin S, put S′=S∖{ae}S' = S \setminus\{a_e\}: then aea_e is covered through ueu_e, and beb_e through cec_e, whose neighbour vev_e is in S′S'. If ce∉Sc_e \notin S, put S′=S∖{ae}S' = S \setminus\{a_e\}: then aea_e is covered through ueu_e, and cec_e has the two neighbours beb_e and vev_e in S′S'. The case ae∉Sa_e \notin S is symmetric to the previous one, with S′=S∖{ce}S' = S \setminus\{c_e\}. In each case the vertices outside SS within distance two of the deleted vertex are covered by S′S', so S′S' is 2-star-forming, against the minimality of SS.

Proof of Theorem 1.4. The order and the bipartition of H♯H^\sharp were noted above. By Lemma 5.1 applied to a largest independent set, β2(H♯)≥n+2m+α(H)\beta_2(H^\sharp) \ge n+2m+\alpha(H), and β2≤SF2\beta_2 \le\mathrm{SF}_2 by [4]. Now let SS be a minimal 2-star-forming set of H♯H^\sharp and YY as in Lemma 5.2. By part (a) the set SS contains both of the vertices v,v′v,v' if v∈Yv \in Y and one of them otherwise, and by parts (b) and (c) it contains at most two of ae,be,cea_e,b_e,c_e, and at most one if both ends of ee lie in YY. Writing e(Y)e(Y) for the number of edges of HH with both ends in YY, we get

∣S∣≤n+∣Y∣+2m−e(Y).|S| \le n + |Y| + 2m - e(Y).

The graph H[Y]H[Y] has at least ∣Y∣−e(Y)|Y| - e(Y) components, and one vertex from each component forms an independent set, so ∣Y∣−e(Y)≤α(H)|Y| - e(Y) \le\alpha(H). Hence ∣S∣≤n+2m+α(H)|S| \le n + 2m + \alpha(H), and the three numbers are equal.

The decision problem lies in NP, since by Lemma 2.1 one checks in polynomial time that a given set SS is 2-star-forming and that S∖{s}S \setminus\{s\} is not, for each s∈Ss \in S. For hardness, HH has an independent set with at least tt vertices if and only if H♯H^\sharp has a minimal 2-star-forming set with at least n+2m+tn + 2m + t vertices, and H♯H^\sharp is built from HH in polynomial time. Node Cover, and with it Independent Set, is NP-complete [9], also on planar graphs [8], and H#H^{\#} is planar when HH is.

For k≥3k \ge3 we have not determined the complexity of computing SFk\mathrm{SF}_k on bipartite graphs.

Remarks

Remark 6.1. For k=2k = 2 the two numbers in Lemmas 3.1 and 3.2 coincide, and indeed β2(G2)=SF2(G2)=6\beta_2(G_2) = \mathrm{SF}_2(G_2) = 6. An exhaustive computation over all subsets shows that SFk(Gk)=3k\mathrm{SF}_k(G_k) = 3k exactly for 3≤k≤73 \le k \le7, so in this range SFk(Gk)/βk(Gk)=3k/(2k+2)\mathrm{SF}_k(G_k)/\beta_k(G_k) = 3k/(2k + 2), which tends to 3/23/2.

Remark 6.2. We computed βk\beta_k and SFk\mathrm{SF}_k for every bipartite graph with at most 12 vertices, generated with nauty [11], and 2≤k≤72 \le k \le7. The smallest bipartite graphs with βk<SFk\beta_k < \mathrm{SF}_k have 10 vertices for k=2k = 2 (the graph of [1], which is unique), 12 vertices for k=3k = 3 (there are 51 such graphs, G3G_3 among them), 10 vertices for k=4k = 4 (a single graph) and 12 vertices for k=5k = 5; for k=6,7k = 6, 7 there is none with at most 12 vertices. Among triangle-free graphs the smallest ones with β3<SF3\beta_3 < \mathrm{SF}_3 have 10 vertices, and there are two of them. We also found no tree with at most 16 vertices and βk<SFk\beta_k < \mathrm{SF}_k for 2≤k≤62 \le k \le6, and, in accordance with Theorem 1.3, no such chain graph with at most 16 vertices for 3≤k≤83 \le k \le8. These computations are not part of the proofs.

Remark 6.3. We also checked Theorem 1.4 by computer in two independent ways: by exhaustive search over all vertex subsets of H#H^{\#} for every connected graph HH with 2n+3m≤262n + 3m \le26, which also confirms parts (a)–(c) of Lemma 5.2 for every minimal 2-star-forming set, and with a SAT solver for every connected graph HH with at most 7 vertices.

Formal verification

Theorems 1.2, 1.3 and 1.4 are formalised in Lean 4 [12] with Mathlib [10]. The definitions follow [13]: a set is kk-star-forming if every outside vertex vv lies in a copy of K1,kK_{1,k} inside G[S∪{v}]G[S \cup\{v\}], minimality is required against every proper subset, and βk\beta_k and SFk\mathrm{SF}_k are the largest sizes of kk-independent and of minimal kk-star-forming sets. For every k≥3k \ge3 the formal statement gives a 2-colouring of GkG_k together with βk(Gk)=2k+2\beta_k(G_k) = 2k + 2 and SFk(Gk)≥3k\mathrm{SF}_k(G_k) \ge3k. Theorem 1.3 is proved for every finite graph with a bipartition in which the neighbourhoods on one side are nested, and the inequality βk≤SFk\beta_k \le\mathrm{SF}_k of [4] is proved along the way. The identity of Theorem 1.4 is proved for every finite graph HH whose vertices are linearly ordered, the order fixing the ends ue<veu_e < v_e, together with a 2-colouring of H♯H^\sharp; here α(H)\alpha(H) is the independence number of Mathlib. The NP-completeness, which follows from the identity by the argument above, is not formalised. The proofs are checked by the kernel and use only the standard axioms of Lean. The development and the search programs are archived in [2].

Declaration of competing interest

The author has no competing interests to declare.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The Lean 4 development, the C and Python checks and the search programs are available at https://github.com/creelie/Conjecture-002 and on Zenodo, doi:10.5281/zenodo.23360032.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this work the author used Claude (Anthropic) to find the construction and the reduction, to check them in C, Python and Lean, and to draft the text. After using this tool, the author reviewed and edited the content as needed and takes full responsibility for the content of the published article.

Author

Deep Bhattacharjee (sole author); [email protected], [email protected]; ORCID 0000-0003-0466-750X.

References

  1. [1]AlyciaBHZ, Refute Sahbi Conjectures 4.1, 4.2 and 7.2 (arXiv:2610.03785): a 10-vertex bipartite graph with dissociation number 5 and a minimal 2-star-forming set of size 6, pull request 14727 to the-omega-institute/trureturning, GitHub, 9 October 2026. https://github.com/the-omega-institute/trureturning/pull/14727
  2. [2]D. Bhattacharjee, Upper k-star-forming sets in bipartite graphs: paper, Lean 4 development and search programs, Zenodo, 2026. doi:10.5281/zenodo.23360032; https://github.com/creelie/Conjecture-002
  3. [3]M. Blidia, M. Chellali, O. Favaron and N. Meddah, On k-independence in graphs with emphasis on trees, Discrete Math. 307 (2007), 2209–2216. doi:10.1016/j.disc.2006.11.007DOI
  4. [4]M. Chellali and O. Favaron, On k-star-forming sets in graphs, J. Combin. Math. Combin. Comput. 68 (2009), 205–214. Zbl 1176.05052.
  5. [5]M. Chellali, O. Favaron, A. Hansberg and L. Volkmann, k-domination and k-independence in graphs: a survey, Graphs Combin. 28 (2012), 1–55. doi:10.1007/s00373-011-1040-3DOI
  6. [6]E. J. Cockayne, O. Favaron, C. Payan and A. G. Thomason, Contributions to the theory of domination, independence and irredundance in graphs, Discrete Math. 33 (1981), 249–258. doi:10.1016/0012-365X(81)90268-5DOI
  7. [7]J. F. Fink and M. S. Jacobson, n-domination in graphs, in: Graph Theory with Applications to Algorithms and Computer Science (Kalamazoo, 1984), Wiley, New York, 1985, 283–300. Zbl 0573.05049.
  8. [8]M. R. Garey, D. S. Johnson and L. Stockmeyer, Some simplified NP-complete graph problems, Theoret. Comput. Sci. 1 (1976), 237–267. doi:10.1016/0304-3975(76)90059-1DOI
  9. [9]R. M. Karp, Reducibility among combinatorial problems, in: Complexity of Computer Computations, Plenum, New York, 1972, 85–103. doi:10.1007/978-1-4684-2001-2_9DOI
  10. [10]The mathlib Community, The Lean mathematical library, in: Proceedings of the 9th ACM SIGPLAN International Conference on Certified Programs and Proofs (CPP 2020), ACM, New York, 2020, 367–381. doi:10.1145/3372885.3378824
  11. [11]B. D. McKay and A. Piperno, Practical graph isomorphism, II, J. Symbolic Comput. 60 (2014), 94–112. doi:10.1016/j.jsc.2013.09.003DOI
  12. [12]L. de Moura and S. Ullrich, The Lean 𝒜 theorem prover and programming language, in: Automated Deduction – CADE 28, Lecture Notes in Comput. Sci. 12699, Springer, Cham, 2021, 625–635. doi:10.1007/978-3-030-79876-5_37
  13. [13]R. Sahbi, Upper k-star-forming sets, k-independence, and upper domination, preprint, arXiv:2610.03785v2, 6 October 2026; withdrawn by the author on 12 October 2026.arxiv.org/abs/2610.03785
  14. [14]M. Yannakakis, Node-deletion problems on bipartite graphs, SIAM J. Comput. 10 (1981), 310–327. doi:10.1137/0210022DOI

Paper details

vlnc:W_kjcut6fn
v1 · latestHistory (1)

published · Fresh — added to the Hub within the last 30 days

added ·

Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

CC BY 4.0

Original source

  • Mathematics

Anthropic Claude Code Opus 5.5 Ultracode · View statement →

Looking for related papers…
Contents