
Due to the advent of the expressions of data other than tabular formats, the topological compositions which make samples interrelated came into prominence. Analogically, those networks can be interpreted as social connections, dataflow maps, citation influence graphs, protein bindings, etc. However, in the case of social networks, it is highly crucial to evaluate the labels of discrete communities. The reason for such a study is the importance of analyzing graph networks to partition the vertices by only using the topological features of network graphs. For each interaction-based entity, a social graph, a mailing dataset, and two citation sets are selected as the testbench repositories. The research mainly focused on evaluating the significance of three artificial intelligence approaches on four different datasets consisting of vertices and edges. Overall, one of these methods so-called "harmonic functions", resulted in the best form to classify those constituents of graph-shaped datasets. This research not only accessed the most valuable method but also determined how graph neural networks work and the need to improve against non-neural network approaches which are faster and computationally cost-effective. Also in this paper, we will show that there is a limit to be accessed by prospective graph neural network variations by using the topological features of trialed networks.
Citation: Hacı İsmail Aslan, Hoon Ko, Chang Choi. Classification of vertices on social networks by multiple approaches[J]. Mathematical Biosciences and Engineering, 2022, 19(12): 12146-12159. doi: 10.3934/mbe.2022565
[1] | Fouad Alkilani, Ouassini Nemraoui, Fareed Ismail . Performance evaluation of solar still integrated with thermoelectric heat pump system. AIMS Energy, 2023, 11(1): 47-63. doi: 10.3934/energy.2023003 |
[2] | Hind I. El-Refaey, Saleh. M. Shalaby, Swellam W. Sharshir, Hewida Omara, Tamer A. Gado . Experimental investigation of the effect of TDS on the thermal performance of the different types of solar stills. AIMS Energy, 2025, 13(2): 428-448. doi: 10.3934/energy.2025016 |
[3] | D.M. Reddy Prasad, R. Senthilkumar, Govindarajan Lakshmanarao, Saravanakumar Krishnan, B.S. Naveen Prasad . A critical review on thermal energy storage materials and systems for solar applications. AIMS Energy, 2019, 7(4): 507-526. doi: 10.3934/energy.2019.4.507 |
[4] | Khadim Ndiaye, Stéphane Ginestet, Martin Cyr . Thermal energy storage based on cementitious materials: A review. AIMS Energy, 2018, 6(1): 97-120. doi: 10.3934/energy.2018.1.97 |
[5] | J. Appelbaum . A static multiple detector solar radiation sensor. AIMS Energy, 2020, 8(5): 802-818. doi: 10.3934/energy.2020.5.802 |
[6] | Khaled M. Bataineh, Assem N. AL-Karasneh . Direct solar steam generation inside evacuated tube absorber. AIMS Energy, 2016, 4(6): 921-935. doi: 10.3934/energy.2016.6.921 |
[7] | Enebe GC, Ukoba K, Jen T-C . Numerical modeling of effect of annealing on nanostructured CuO/TiO2 pn heterojunction solar cells using SCAPS. AIMS Energy, 2019, 7(4): 527-538. doi: 10.3934/energy.2019.4.527 |
[8] | Kaovinath Appalasamy, R Mamat, Sudhakar Kumarasamy . Smart thermal management of photovoltaic systems: Innovative strategies. AIMS Energy, 2025, 13(2): 309-353. doi: 10.3934/energy.2025013 |
[9] | Thomas Wetzel, Julio Pacio, Luca Marocco, Alfons Weisenburger, Annette Heinzel, Wolfgang Hering, Carsten Schroer, Georg Muller, Jurgen Konys, Robert Stieglitz, Joachim Fuchs, Joachim Knebel, Concetta Fazio, Markus Daubner, Frank Fellmoser . Liquid metal technology for concentrated solar power systems: Contributions by the German research program. AIMS Energy, 2014, 2(1): 89-98. doi: 10.3934/energy.2014.1.89 |
[10] | Chao-Jen Li, Peiwen Li, Kai Wang, Edgar Emir Molina . Survey of Properties of Key Single and Mixture Halide Salts for Potential Application as High Temperature Heat Transfer Fluids for Concentrated Solar Thermal Power Systems. AIMS Energy, 2014, 2(2): 133-157. doi: 10.3934/energy.2014.2.133 |
Due to the advent of the expressions of data other than tabular formats, the topological compositions which make samples interrelated came into prominence. Analogically, those networks can be interpreted as social connections, dataflow maps, citation influence graphs, protein bindings, etc. However, in the case of social networks, it is highly crucial to evaluate the labels of discrete communities. The reason for such a study is the importance of analyzing graph networks to partition the vertices by only using the topological features of network graphs. For each interaction-based entity, a social graph, a mailing dataset, and two citation sets are selected as the testbench repositories. The research mainly focused on evaluating the significance of three artificial intelligence approaches on four different datasets consisting of vertices and edges. Overall, one of these methods so-called "harmonic functions", resulted in the best form to classify those constituents of graph-shaped datasets. This research not only accessed the most valuable method but also determined how graph neural networks work and the need to improve against non-neural network approaches which are faster and computationally cost-effective. Also in this paper, we will show that there is a limit to be accessed by prospective graph neural network variations by using the topological features of trialed networks.
A topological descriptor is a single number that represents a chemical structure in graph-theoretical terms via the molecular graph, they play a significant role in mathematical chemistry, especially in the QSPR/QSAR investigations. A topological descriptor is called a topological index if it correlates with a molecular property.
Topological indices are used to understand physico-chemical properties of chemical compounds, since they capture the essence of some properties of a molecule in a single number. Several of the well-known topological indices based its computing as a sum of terms which correspond to edges where each term is obtained by a symmetric expression of the degree of both end points, e.g., harmonic, Randić connectivity, first and second Zagreb, atom-bound connectivity and Platt indices. Lately, researchers in chemistry and pharmacology have focused in topological indices based on degrees of vertices obtaining good results and showing that a number of these indices has turned out to be a useful tool, see [1] and references therein. Probably, Randić connectivity index (R) [2] is the most known. In fact, there exist hundreds of works about this molecular descriptor (see, e.g., [3,4,5,6,7,8] and the references therein). Trying to improve the predictive power of this index during many years, scientists proposed a great number of new topological descriptors based on degrees of vertices, similar to the Randić index. There are a lot of works showing the interest on these indices, see e.g., [9,10,11,12,13,14,15,16,17].
The study of the exponential vertex-degree-based topological indices was initiated in [18] and has been successfully studied in [19,20,21,22,23,24,25]. The study of topological indices associated to exponential representations has been successfully studied in [12,14,20,22,26,27]. Cruz et al. mentioned in 2020 some open problems on the exponential vertex-degree-based topological indices of trees [28] and Das et al. show in 2021 the solution of two of those problems in [23]. In this sense, we work with generic exponential topological indices described below that allows us to improve some of the bounds given in [23] as well as obtain some new results.
Along this work, given a graph G=(V(G),E(G)) and a symmetric function f:[1,∞)×[1,∞)→R, we consider the general topological indices
F(G)=∑uv∈E(G)f(du,dv),eF(G)=∑uv∈E(G)ef(du,dv). |
As usual, uv denotes an edge of the graph G connecting the vertices u and v, and dx denotes the degree of the vertex x. The family of the topological indices like the one to the right above are called exponential for obvious reasons.
Two examples of those topological indices that have been extended to its exponential indices are the generalized atom-bound connectivity index, ABCα, and the generalized second Zagreb index, Mα2, defined respectively as
ABCα(G)=∑uv∈E(G)(du+dv−2dudv)α, |
Mα2(G)=∑uv∈E(G)(dudv)α. |
Then, their corresponding exponential indices are
eABCα(G)=∑uv∈E(G)e(du+dv−2dudv)α, |
eMα2(G)=∑uv∈E(G)e(dudv)α. |
The study of topological indices is often formalized mathematically as optimization problems on graphs. In general, those problems have been proven to be useful and quite difficult. Indeed, to obtain quasi-minimizing or quasi-maximizing graphs is a good strategy that is commonly used. Therefore, frequently, it is needed to find bounds on them which involve several parameters.
Topological indices have been successfully applied to different branches of knowledge, such as chemistry, physics, biology, social sciences, etc. see [29,30,31,32,33]. In this direction, two of the most important theoretical and practical problems facing the study of topological indices are the following: the study of the extremal problems associated with topological indices and obtaining new inequalities relating different indices. This research has as fundamental contribution to face the above mentioned problems associated with the study of exponential vertex-degree-based topological indices. In particular, we solve some extremal problems related with the exponential vertex-degree-based topological indices eMα2 and eABCα. Also, we obtain new lower and upper optimal bounds for these general exponential indices.
Throughout this work, G=(V(G),E(G)) denotes a (non-oriented) finite simple graph without isolated vertices. Given a graph G and v∈V(G), we denote by N(v) the set of neighbors of v, i.e., N(v)={u∈V(G)|uv∈E(G)}. We denote by Δ,δ,n,m the maximum degree, the minimum degree and the cardinality of the set of vertices and edges of G, respectively; thus, 1≤δ≤Δ<n.
We study in this section some optimization problems on the exponential index eMα2.
Recall that we denote by N(w) the set of neighbors of w.
Proposition 2.1. Let G be a graph with nonadjacent vertices u and v. If f is a symmetric function which is increasing in each variable, then eF(G+uv)>eF(G).
Proof. Since f is increasing in each variable, we have ef(du+1,dw)≥ef(du,dw), ef(dv+1,dw)≥ef(dv,dw) and
eF(G+uv)−eF(G)=∑w∈N(u)(ef(du+1,dw)−ef(du,dw))+∑w∈N(v)(ef(dv+1,dw)−ef(dv,dw))+ef(du+1,dv+1)≥ef(du+1,dv+1)>0. |
Hence,
eF(G+uv)>eF(G). |
For example, if we take α>0, f(x,y)=(xy)α, the cycle Cn with n>2 and remove one edge we obtain the path Pn, and we have eMa2(Cn)=ne4α, eMa2(Pn)=(n−3)e4α+2e2α and eMa2(Cn)−eMa2(Pn)=3e4α−2e2α>0. Now if we take the complete graph Kn with n>2 and H is the graph obtained by removing an edge from Kn, we have eMa2(Kn)=n(n−1)2e(n−1)2α, eMa2(H)=((n−2)(n−3)2)e(n−1)2α+2(n−2)e[(n−1)(n−2)]α and eMa2(Kn)−eMa2(H)=(2n−3)e(n−1)2α−2(n−2)e[(n−1)(n−2)]α>0. Thus, we have e(n−1)α−(n−2)α>1>2n−42n−3.
Given an integer n≥2, let G(n) be the set of graphs with n vertices.
Given integers 1≤δ<n, let H(n,δ) be the set of graphs with n vertices and minimum degree δ
We consider the optimization problem for the exponential index eMα2 on G(n).
Theorem 2.2. Consider α>0 and an integer n≥2.
(1) The unique graph that maximizes the eMα2 index on G(n) is the complete graph Kn.
(2) If n is even, then the unique graph that minimizes the eMα2 index on G(n) is the disjoint union of n/2 paths P2. If n is odd, then the unique graph that minimizes the eMα2 index on G(n) is the disjoint union of (n−3)/2 paths P2 and a path P3.
Proof. Since α>0 we have that f(x,y)=(xy)α is an increasing function in each variable and so, we can apply Proposition 2.1. The first item is a direct consequence of Proposition 2.1.
Assume first that n is even. Handshaking lemma gives 2m≥nδ≥n. We have for any graph G∈G(n)
eMα2(G)=∑uv∈E(G)e(dudv)α≥∑uv∈E(G)e=me≥12ne, |
and the equality in the bound is attained if and only if du=1 for every u∈V(G), i.e., G is the disjoint union of n/2 path graphs P2.
Assume now that n is odd, and consider a graph G∈G(n). If du=1 for every u∈V(G), then handshaking lemma gives 2m=n, a contradiction since n is odd. Thus, there exists a vertex w with dw≥2. Handshaking lemma gives 2m≥(n−1)δ+2≥n+1. We have
eMα2(G)=∑u∈N(w)e(dudw)α+∑uv∈E(G),u,v≠we(dudv)α≥∑u∈N(w)e2α+∑uv∈E(G),u,v≠we≥2e2α+(m−2)e≥2e2α+(n+12−2)e=2e2α+n−32e, |
and the equality in the bound is attained if and only if du=1 for every u∈V(G)∖{w}, and dw=2, i.e., G is the disjoint union of (n−3)/2 path graphs P2 and a path graph P3.
Note that for α=1, the result in Theorem 2.2 was obtained in [20,Theorem 2.2].
If 1≤δ<Δ are integers, we say that a graph G is (Δ,δ)-pseudo-regular if there exists v∈V(G) with dv=Δ and du=δ for every u∈V(G)∖{v}.
In [34] appears the following result.
Lemma 2.3. Consider integers 2≤k<n.
(1) If kn is even, then there is a connected k-regular graph with n vertices.
(2) If kn is odd, then there is a connected (k+1,k)-pseudo-regular graph with n vertices.
Given integers 1≤δ<n, denote by Kδn the n-vertex graph with maximum degree n−1 and minimum degree δ, obtained from the complete graph Kn−1 and an additional vertex v in the following way: Fix δ vertices u1,…,uδ∈V(Kn−1) and let V(Kδn)=V(Kn−1)∪{v} and E(Kδn)=E(Kn−1)∪{u1v,…,uδv}.
Figure 1 illustrates this construction by showing the graphs K26 and K36.
We consider now the optimization problem for the exponential index eMα2 on H(n,δ).
Theorem 2.4. Consider α>0 and integers 1≤δ<n.
(1) Then the unique graph in H(n,δ) that maximizes the eMα2 index is Kδn.
(2) If δ≥2 and δn is even, then the unique graphs in H(n,δ) that minimize the eMα2 index are the δ-regular graphs.
(3) If δ≥2 and δn is odd, then the unique graphs in H(n,δ) that minimize the eMα2 index are the (δ+1,δ)-pseudo-regular graphs.
Proof. Given a graph G∈H(n,δ)∖{Kδn}, fix any vertex u∈V(G) with du=δ. Since
G≠G∪{vw:v,w∈V(G)∖{u} and vw∉E(G)}=Kδn, |
Proposition 2.1 gives eMα2(Kδn)>eMα2(G). This proves item (1).
Handshaking lemma gives 2m≥nδ.
Since du≥δ for every u∈V(G), we obtain
eMα2(G)=∑uv∈E(G)e(dudv)α≥∑uv∈E(G)eδ2α=meδ2α≥12nδeδ2α, |
and the equality in the bound is attained if and only if du=δ for every u∈V(G).
If δn is even, then Lemma 2.3 gives that there is a δ-regular graph with n vertices. Hence, the unique graphs in H(n,δ) that minimize the eMα2 index are the δ-regular graphs.
If δn is odd, then handshaking lemma gives that there is no regular graph. Hence, there exists a vertex w with dw≥δ+1. Since du≥δ for every u∈V(G), handshaking lemma gives 2m≥(n−1)δ+δ+1=nδ+1. We have
eMα2(G)=∑u∈N(w)e(dudw)α+∑uv∈E(G),u,v≠we(dudv)α≥∑u∈N(w)eδα(δ+1)α+∑uv∈E(G),u,v≠weδ2α. |
From the above and since w has at least δ+1 neighbors we have
eMα2(G)≥(δ+1)eδα(δ+1)α+(m−δ−1)eδ2α, |
now using 2m≥nδ+1, we have
eMα2(G)≥(δ+1)eδα(δ+1)α+(nδ+12−δ−1)eδ2α=(δ+1)eδα(δ+1)α+12(δ(n−2)−1)eδ2α, |
and equality in the bound is attained if and only if du=δ for every u∈V(G)∖{w}, and dw=δ+1. Lemma 2.3 gives that there is a (δ+1,δ)-pseudo-regular graph with n vertices. Therefore, the unique graphs in H(n,δ) that minimize the eMα2 index are the (δ+1,δ)-pseudo-regular graphs.
In [23,Theorem 2.1] appears the inequality
eABC(G)≥Δ(e√2(Δ−1)Δ−e√1−1Δ)+meABC(G)m. |
The equality holds in this bound if and only if G is a disjoint union of isolated edges or each connected component of G is a path graph Pk (k≥3) or a cycle graph Ck (k≥3).
Proposition 3.1. If G is a graph with size m, then
eABC(G)≥meABC(G)m. |
The equality holds if every edge of G is incident to a vertex of degree 2, or G is regular or biregular.
Proof. ABC(G)=∑uv∈E(G)(du+dv−2dudv). Note that the exponential function exp(x)=ex is a strictly convex function, and Jensen's inequality gives
exp(1m∑uv∈E(G)du+dv−2dudv)≤1m∑uv∈E(G)edu+dv−2dudv, |
and the equality in this bound is attained if and only if du+dv−2dudv=dw+dz−2dwdz for every uv,wz∈E(G).
If every edge of G is incident to a vertex of degree 2, then du+dv−2dudv=12 for every uv∈E(G). If G is a regular or biregular graph, then du+dv−2dudv=Δ+δ−2Δδ for every uv∈E(G).
Since √2(Δ−1)Δ<√1−1Δ if Δ>2, Propostion 3.1 improves [23,Theorem 2.1].
The following result relates the exponential generalized atom-bound connectivity indices with positive and negative parameters.
Theorem 3.2. Let G be a graph with m edges, minimum degree δ, maximum degree Δ>2, and without isolated edges, and let α,β∈R with α>0>β. Then
e((2(Δ−1)Δ2)α−(2(Δ−1)Δ2)β)eABCβ(G)≤eABCα(G), |
and the equality holds if and only if G is regular.
If δ=1, then
eABCα(G)≤e((1−1Δ)α−(1−1Δ)β)eABCβ(G), |
and the equality holds if and only if G is the disjoint union of star graphs K1,Δ.
If δ≥2, then
eABCα(G)≤e((2(δ−1)δ2)α−(2(δ−1)δ2)β)eABCβ(G). |
If δ>2, then the equality holds if and only if G is regular. If δ=2, then the equality holds if and only if each edge in G is incident to a vertex with degree two.
Proof. For each fixed α and β with α>0>β, we are going to compute the extremal values of the function g:[δ,Δ]×[max{2,δ},Δ] (with Δ≥3) given by
g(x,y)=e(x+y−2xy)αe−(x+y−2xy)β=e((x+y−2xy)α−(x+y−2xy)β). |
We have
∂g∂x=e((x+y−2xy)α−(x+y−2xy)β)(α(x+y−2xy)α−1−β(x+y−2xy)β−1)1x2(2y−1)≤0. |
Then the function g(x,y) is decreasing on x∈[δ,Δ] for each fixed y∈[max{2,δ},Δ] and consequently
g(Δ,y)≤g(x,y)≤g(δ,y). |
Let us define
g1(y)=g(Δ,y)=e((y+Δ−2Δy )α−(y+Δ−2Δy )β). |
We have
g′1(y)=e((y+Δ−2Δy )α−(y+Δ−2Δy )β)×(α(y+Δ−2Δy)α−1−β(y+Δ−2Δy)β−1)1y2(2Δ−1)<0. |
Then g1(y) is strictly decreasing, and consequently g1(Δ)≤g1(y). Therefore, for each uv∈E(G) we have
e((2(Δ−1)Δ2 )α−(2(Δ−1)Δ2 )β )≤e(du+dv−2dudv )αe−(du+dv−2dudv )β,e((2(Δ−1)Δ2 )α−(2(Δ−1)Δ2 )β )e(du+dv−2dudv )β≤e(du+dv−2dudv )α,e((2(Δ−1)Δ2 )α−(2(Δ−1)Δ2 )β )eABCβ(G)≤eABCα(G), |
and the equality in the last inequality holds if and only if du=dv=Δ for each uv∈E(G), i.e., G is a regular graph.
Let us define
g2(y)=g(δ,y)=e((y+δ−2δy )α−(y+δ−2δy )β). |
We have
g′2(y)=e((y+δ−2δy )α−(y+δ−2δy )β)×(α(y+δ−2δy)α−1−β(y+δ−2δy)β−1)1y2(2δ−1). |
If δ=1, then g′2(y)>0 and so, g2(y) is strictly increasing. Consequently, g2(y)≤g2(Δ), and we have for each uv∈E(G)
e(du+dv−2dudv )αe−(du+dv−2dudv )β≤e((1−1Δ )α−(1−1Δ )β),e(du+dv−2dudv )α≤e((1−1Δ )α−(1−1Δ )β)e(du+dv−2dudv )β,eABCα(G)≤e((1−1Δ )α−(1−1Δ )β)eABCβ(G), |
and the equality in the last inequality holds if and only if du=1 and dv=Δ o viceversa for each uv∈E(G), i.e., G is the disjoint union of star graphs K1,Δ.
If δ=2, then g′2(y)=0 and so, g2(y) is a constant function. We have
g(x,y)≤g(2,y)=g2(y)=e((12 )α−(12 )β)=e((2(δ−1)δ2 )α−(2(δ−1)δ2 )β),e(du+dv−2dudv )αe−(du+dv−2dudv )β≤e((2(δ−1)δ2 )α−(2(δ−1)δ2 )β),e(du+dv−2dudv )α≤e((2(δ−1)δ2 )α−(2(δ−1)δ2 )β)e(du+dv−2dudv )β,eABCα(G)≤e((2(δ−1)δ2 )α−(2(δ−1)δ2 )β)eABCβ(G), |
and the equality in the last inequality holds if and only if each edge in G is incident to a vertex with degree two.
If δ>2, then g′2(y)<0 and so, g2(y) is strictly decreasing. Consequently, g2(y)≤g2(δ), and we have for each uv∈E(G)
e(du+dv−2dudv )αe−(du+dv−2dudv )β≤e((2(δ−1)δ2 )α−(2(δ−1)δ2 )β),e(du+dv−2dudv )α≤e((2(δ−1)δ2 )α−(2(δ−1)δ2 )β)e(du+dv−2dudv )β,eABCα(G)≤e((2(δ−1)δ2 )α−(2(δ−1)δ2 )β)eABCβ(G), |
and the equality in the last inequality holds if and only if du=dv=δ for each uv∈E(G), i.e., G is a regular graph.
The hypothesis Δ>2 in Theorem 3.2 is not an important restriction: If Δ=2, then G is is the disjoint union of path and cycle graphs with three or more vertices, and we have eABCα(G)=me(12)α. Hence,
eABCα(G)=e((12)α−(12)β)eABCβ(G). |
We have studied some properties of the generalized exponential indices. For the exponential index eMα2 with α>0, we characterize the graphs with extreme values in the class of graphs with a fixed number of vertices and in the class of graphs with a fixed minimum degree and a fixed number of vertices.
In addition, we found some optimal inequalities involving the exponential atom-bound connectivity index. In particular, we found a bound that improves a result in [23]. Also, we obtained an inequality concerning the indices eABCα for different values of the parameter.
As an open problem it remains to find the extremal graphs and to obtain optimal bounds for other generalized exponential vertex-degree-based topological indices. In particular, for the index eMa2 to find the extremal graphs in other classes, for example the class of graphs with n vertices and fixed maximum degree.
This research is supported by a grant from Agencia Estatal de Investigación (PID2019-106433GB-I00 / AEI / 10.13039/501100011033), Spain.
The author declare there is no conflict of interest.
[1] |
A. Hogan, E. Blomqvist, M. Cochez, C. D'amato, G. De Melo, C. Gutierrez, et al., Knowledge graphs, ACM Comput. Surv., 54 (2022), 1–37. https://doi.org/10.1145/3447772 doi: 10.1145/3447772
![]() |
[2] |
M. M. Bronstein, J. Bruna, Y. LeCun, A. Szlam, P. Vandergheynst, Geometric deep learning: going beyond euclidean data, IEEE Signal Process Mag., 34 (2017), 18–42. https://doi.org/10.1109/MSP.2017.2693418 doi: 10.1109/MSP.2017.2693418
![]() |
[3] |
Z. Wu, S. Pan, F. Chen, G. Long, C. Zhang, P. S. Yu, A comprehensive survey on graph neural networks, IEEE Trans. Neural Networks Learn. Syst., 32 (2021), 4–24. https://doi.org/10.1109/TNNLS.2020.2978386 doi: 10.1109/TNNLS.2020.2978386
![]() |
[4] |
S. Kearnes, K. McCloskey, M. Berndl, V. Pande, P. Riley, Molecular graph convolutions: moving beyond fingerprints, J. Comput.-Aided Mol. Des., 30 (2016), 595–608. https://doi.org/10.1007/s10822-016-9938-8 doi: 10.1007/s10822-016-9938-8
![]() |
[5] | A. Fout, J. Byrd, B. Shariat, A. Ben-Hur, Protein interface prediction using graph convolutional networks, in Advances in Neural Information Processing Systems, 30 (2017), 6533–6542. Available from: https://proceedings.neurips.cc/paper/2017/file/f507783927f2ec2737ba40afbd17efb5-Paper.pdf. |
[6] | E. Choi, Z. Xu, Y. Li, M. Dusenberry, G. Flores, E. Xue, et al., Learning the graphical structure of electronic health records with graph convolutional transformer, in Proceedings of the AAAI Conference on Artificial Intelligence, 34 (2020), 606–613. https://doi.org/10.1609/aaai.v34i01.5400 |
[7] | M. Zhang, Y. Chen, Link prediction based on graph neural networks, in Advances in Neural Information Processing Systems, 31 (2018), 5171–5181. Available from: https://proceedings.neurips.cc/paper/2018/file/53f0d7c537d99b3824f0f99d62ea2428-Paper.pdf. |
[8] | C. Li, D. Goldwasser, Encoding social information with graph convolutional networks forPolitical perspective detection in news media, in Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, 2019 (2019), 2594–2604. https://doi.org/10.18653/v1/P19-1247 |
[9] | T. Bian, X. Xiao, T. Xu, P. Zhao, W. Huang, Y. Rong, et al., Rumor detection on social media with Bi-directional graph convolutional networks, in Proceedings of the AAAI Conference on Artificial Intelligence, 34 (2020), 549–556. https://doi.org/10.1609/aaai.v34i01.5393 |
[10] | M. Schlichtkrull, T. N. Kipf, P. Bloem, R. van den Berg, I. Titov, M. Welling, Modeling relational data with graph convolutional networks, in The Semantic Web, (2018), 593–607. https://doi.org/10.1007/978-3-319-93417-4_38. |
[11] | N. Park, A. Kan, X. L. Dong, T. Zhao, C. Faloutsos, Estimating node importance in knowledge graphs using graph neural networks, in Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, (2019), 596–606. https://doi.org/10.1145/3292500.3330855 |
[12] |
Z. Cui, K. Henrickson, R. Ke, Y. Wang, Traffic graph convolutional recurrent neural network: a deep learning framework for network-scale traffic learning and forecasting, IEEE Trans. Intell. Transp. Syst., 21 (2020), 4883–4894. https://doi.org/10.1109/TITS.2019.2950416 doi: 10.1109/TITS.2019.2950416
![]() |
[13] | H. Wu, L. Cheng, J. Jin, F. Yuan, Dialog acts classification with semantic and structural information, in 2019 International Conference on Intelligent Computing, Automation and Systems (ICICAS), (2019), 438–442. https://doi.org/10.1109/ICICAS48597.2019.00098 |
[14] | Y. T. Lin, M. T. Wu, K. Y. Su, Syntax-aware natural language inference with graph matching networks, in 2020 International Conference on Technologies and Applications of Artificial Intelligence (TAAI), (2020), 85–90. https://doi.org/10.1109/TAAI51410.2020.00024 |
[15] | Z. Wang, H. Chang, 3D mesh deformation using graph convolution network, in 2019 IEEE 4th International Conference on Computer and Communication Systems (ICCCS), (2019), 375–378. https://doi.org/10.1109/CCOMS.2019.8821790 |
[16] | P. Pradhyumna, G. P. Shreya, Mohana, Graph Neural Network (GNN) in image and video understanding using deep learning for computer vision applications, in 2021 Second International Conference on Electronics and Sustainable Communication Systems (ICESC), (2021), 1183–1189. https://doi.org/10.1109/ICESC51422.2021.9532631 |
[17] |
S. Zhang, H. Tong, J. Xu, R. Maciejewski, Graph convolutional networks: a comprehensive review, Comput. Social Networks, 6 (2019), 11. https://doi.org/10.1186/s40649-019-0069-y doi: 10.1186/s40649-019-0069-y
![]() |
[18] | W. Hamilton, Z. Ying, J. Leskovec, Inductive representation learning on large graphs, in Advances in Neural Information Processing Systems, 30 (2017). Available from: https://proceedings.neurips.cc/paper/2017/file/5dd9db5e033da9c6fb5ba83c7a7ebea9-Paper.pdf. |
[19] | J. Gasteiger, A. Bojchevski, S. Günnemann, Predict then propagate: graph neural networks meet personalized pageRank, preprint, arXiv: 1810.05997. |
[20] | L. Page, S. Brin, R. Motwani, T. Winograd, The pageRank citation ranking: bringing order to the web, 1999. Available from: http://ilpubs.stanford.edu:8090/422/. |
[21] | F. Wu, A. Souza, T. Zhang, C. Fifty, T. Yu, K. Weinberger, Simplifying graph convolutional networks, in Proceedings of the 36th International Conference on Machine Learning, 97 (2019), 6861–6871. Available from: https://proceedings.mlr.press/v97/wu19e.html. |
[22] | P. Veličković, G. Cucurull, A. Casanova, A. Romero, P. Liò, Y. Bengio, Graph attention networks, preprint, arXiv: 1710.10903. |
[23] | P. Veličković, W. Fedus, W. L. Hamilton, P. Liò, Y. Bengio, R. D. Hjelm, Deep graph infomax, preprint, arXiv: 1809.10341. |
[24] | G. Nikolentzos, M. Vazirgiannis, Random walk graph neural networks, in Advances in Neural Information Processing Systems, 33 (2020), 16211–16222. Available from: https://proceedings.neurips.cc/paper/2020/file/ba95d78a7c942571185308775a97a3a0-Paper.pdf. |
[25] | B. Nica, A brief introduction to spectral graph theory, 2016. Available from: https://arXiv.org/pdf/1609.08072.pdf. |
[26] |
I. Benjamini, L. Lovász, Harmonic and analytic functions on graphs, J. Geom., 76 (2003), 3–15. https://doi.org/10.1007/s00022-033-1697-8 doi: 10.1007/s00022-033-1697-8
![]() |
[27] |
W. W. Zachary, An information flow model for conflict and fission in small groups, J. Anthropol. Res., 33 (1977), 452–473. https://doi.org/10.1086/jar.33.4.3629752 doi: 10.1086/jar.33.4.3629752
![]() |
[28] | J. Leskovec, A. Krevl, SNAP datasets: Stanford Large Network Dataset Collection, 2014. Available from: http://snap.stanford.edu/data. |
[29] |
A. Bharali, An analysis of Email-Eu-Core network, Int. J. Sci. Res. Math. Stat. Sci., 5 (2018), 100–104. https://doi.org/10.26438/ijsrmss/v5i4.100104 doi: 10.26438/ijsrmss/v5i4.100104
![]() |
[30] |
D. Grattarola, C. Alippi, Graph neural networks in TensorFlow and Keras with Spektral, IEEE Comput. Intell. Mag., 16 (2021), 99–106. https://doi.org/10.1109/MCI.2020.3039072 doi: 10.1109/MCI.2020.3039072
![]() |
[31] | T. N. Kipf, M. Welling, Semi-supervised classification with graph convolutional networks, preprint, arXiv: 1609.02907. |
[32] |
F. Scarselli, M. Gori, A. C. Tsoi, M. Hagenbuchner, G. Monfardini, The graph neural network model, IEEE Trans. Neural Networks, 20 (2009), 61–80. https://doi.org/10.1109/TNN.2008.2005605 doi: 10.1109/TNN.2008.2005605
![]() |
[33] | X. Zhu, Z. Ghahramani, J. D. Lafferty, Semi-supervised learning using Gaussian fields and harmonic functions, in Proceedings of the Twentieth International Conference on Machine Learning, (2003), 912–919. Available from: https://dl.acm.org/doi/10.5555/3041838.3041953. |
[34] | A. Hagberg, D. S. Chult, P. J. Swart, Exploring network structure, dynamics, and function using networkx, in Proceedings of the 7th Python in Science conference (SciPy 2008), (2008), 11–15. Available from: https://conference.scipy.org/proceedings/scipy2008/paper_2/. |
[35] | L. He, C. T. Lu, J. Ma, J. Cao, L. Shen, P. S. Yu, Joint community and structural hole spanner detection via harmonic modularity, in Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, (2016), 875–884. https://doi.org/10.1145/2939672.2939807 |
[36] | S. Luan, C. Hua, Q. Lu, J. Zhu, M. Zhao, S. Zhang, et al., Is heterophily a real nightmare for graph neural networks to do node classification? preprint, arXiv: 2109.05641. |
[37] |
Q. Yao, R. Y. M. Li, L. Song, Construction safety knowledge sharing on YouTube from 2007 to 2021: Two-step flow theory and semantic analysis, Saf. Sci., 153 (2022), 105796. https://doi.org/10.1016/j.ssci.2022.105796 doi: 10.1016/j.ssci.2022.105796
![]() |
[38] |
Q. Yao, R. Y. M. Li, L. Song, M. J. C. Crabbe, Construction safety knowledge sharing on Twitter: a social network analysis, Saf. Sci., 143 (2021), 105411. https://doi.org/10.1016/j.ssci.2021.105411 doi: 10.1016/j.ssci.2021.105411
![]() |
1. | Fouad Alkilani, Ouassini Nemraoui, Fareed Ismail, Performance evaluation of solar still integrated with thermoelectric heat pump system, 2023, 11, 2333-8334, 47, 10.3934/energy.2023003 | |
2. | S. Balakrishnan, R. Harish, R. Mukesh, K. Rajkumar, R. Seenivasan, N. Nwulu, K. Natarajan, R. Shanmugasundaram, P.K. Balachandran, M. Krishnamoorthy, P. Mounica, Numerical Investigation of Solar Still Desalination Efficiency across Different Salinity Levels, 2024, 547, 2267-1242, 01015, 10.1051/e3sconf/202454701015 | |
3. | Hossam AbdelMeguid, Waleed M. El Awady, Theoretical investigation into saline optical properties for enhancing solar still performance: Mathematical modeling approach, 2024, 54, 24519049, 102846, 10.1016/j.tsep.2024.102846 | |
4. | Mohammed El Hadi Attia, Abd Elnaby Kabeel, M.A. Elazab, Mahmoud Bady, Effect of charcoal balls on distillation productivity in hemispherical solar stills: Experimental investigation and performance analysis, 2024, 282, 0038092X, 112954, 10.1016/j.solener.2024.112954 | |
5. | A.S. Abdullah, Fadl A. Essa, Hitesh Panchal, Wissam H. Alawee, Ammar H. Elsheikh, Enhancing the performance of tubular solar stills for water purification: A comprehensive review and comparative analysis of methodologies and materials, 2024, 21, 25901230, 101722, 10.1016/j.rineng.2023.101722 | |
6. | Faiz T Jodah, Wissam H Alawee, Hayder A Dhahad, ZM Omara, Evaluating the performance of spherical, hemispherical, and tubular solar stills with various configurations - A detailed review, 2024, 238, 0957-6509, 1261, 10.1177/09576509241266284 | |
7. | Omar Bait, A critical review on triangular pyramid, weir–type, spherical, and hemispherical solar water distiller conceptions, 2024, 269, 0038092X, 112322, 10.1016/j.solener.2024.112322 | |
8. | Muntadher Mohammed Ali Saeed, Hassanain Ghani Hameed, Assaad A. Abbass, Design and Testing of a Single‐Slope Solar Still With a Semi‐Elliptical Absorbent Base: An Experimental and Numerical Study, 2025, 2688-4534, 10.1002/htj.23258 | |
9. | Ahmed Kadhim Hussein, Farhan Lafta Rashid, Mohammed Kawa Rasul, Ali Basem, Obai Younis, Raad Z. Homod, Mohammed El Hadi Attia, Mudhar Al-Obaidi, Zainab T Al-Sharify, Noor M. Al-Dabooni, Bagh Ali, Sachindra Kumar Rout, Enhancing hemispherical solar stills thermal performance by using nanotechnologies – recent advances and overview, 2025, 14, 2162-2515, 127, 10.1080/21622515.2025.2458093 | |
10. | Fadl A. Essa, Bahaa Saleh, Abdullah A. Algethami, Sunday O. Oyedepo, Z.M. Omara, Mahmoud S. El-Sebaey, Enhanced solar desalination via hemispheric distiller with thermal storage, heaters, and condensation: Exergoeconomic and environmental analysis, 2025, 285, 09270248, 113529, 10.1016/j.solmat.2025.113529 | |
11. | Zahraa Hamzah Hasan, Ammar Muslim Hadi, Wisam A. Abd Al‐wahid, Dhafer Manea Hachim, Numerical Investigation of Hemispherical Solar Still Performance With Optimal Amount of PCM as a Heat Storage, 2025, 2688-4534, 10.1002/htj.23335 | |
12. | Hossam AbdelMeguid, Multi-stage PCM-enhanced solar distiller desalination: investigating energy storage dynamics, phase transition behavior, and system performance with variable PCM/saline mass ratios, 2025, 122, 2352152X, 116687, 10.1016/j.est.2025.116687 | |
13. | Ramnarayan Meena, Hashim Sahar Mohaisen, Lav Kumar Kaushik, Vivek Patel, Vinay Pandit, Yogeshkumar Khimsuriya, Pranav Mehta, Ravishankar Sathyamurthy, Abd Elnaby Kabeel, Performance enhancement of pyramid solar still integrated with evacuated tube collector, phase change material, and fins: Thermal, exergy, economic, and environmental analysis, 2025, 127, 2352152X, 117094, 10.1016/j.est.2025.117094 |