Pith. sign in

REVIEW 3 major objections 6 minor 54 references

IMMENSE: Inductive Multi-perspective User Classification in Social Networks

T0 review · 3 major / 6 minor · reviewed 2026-08-08 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read IMMENSE is an inductive hybrid classifier that identifies risky social-media users by fusing content semantics, social topology, and spatial closeness, and it is reported to outperform five state-of-the-art competitors even when the…

desk verdict Useful inductive multi-perspective framing, but the evaluation labels are generated from the method's own features, so the superiority claim is unsupported as stated. read the letter →

arxiv 2608.05259 v1 pith:BXB76RCK submitted 2026-08-05 cs.SI cs.AI

Francesco Benedetti , Antonio Pellicani , Gianvito Pio , Michelangelo Ceci This is my paper · ORCID
classification cs.SIcs.AI
keywords riskyuseridentificationinductivelearningsocialnetworkanalysistextembeddingsgraphneuralnetworksspatialclosenessclassimbalancemulti-perspectiveclassification
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper introduces IMMENSE, a machine-learning method for flagging users who spread hateful, discriminatory, or violent content on social networks. Its central claim is that a hybrid model combining three perspectives—the semantics of a user's posts, their follow relationships, and their geographic closeness to other users—classifies risky users more accurately than any single perspective, and does so inductively: once trained, it can label users it has never seen without retraining. On a Twitter/X dataset of 37,945 users with roughly 7% risky users, IMMENSE is reported to outperform five competitors, including SAIRUS, even though SAIRUS is run in a transductive setting that lets it see test users during training. If the result holds, law-enforcement agencies get a reusable tool that can score a batch of new users in seconds rather than retraining a model from scratch.

What carries the argument

The machinery is a per-perspective encoder plus a fusion layer. The text module exploits the additive compositionality of word embeddings to turn a user's chronological post stream into one vector, and uses stacked autoencoders to produce per-class reconstruction errors. The relational and spatial modules both use GraphSAGE—the first on the directed follow graph initialized with the semantic embeddings, the second on a graph whose edge weights are normalized geographic closeness—to output safe/risky probabilities. An MLP with one hidden layer learns to combine the three modules' seven outputs. The inductive property comes from GraphSAGE and the autoencoders operating on local information: a new user's representation depends only on their own posts and immediate neighborhood, so the trained model can be applied to unseen nodes or a whole new network without retraining.

What would settle it

Re-run the same comparison on a dataset whose risky/safe labels come from outside the labeling rule, for example expert-annotated users or accounts later suspended by the platform for harmful content, and check whether IMMENSE's F1 advantage over the five competitors persists; if the margin vanishes or reverses, the reported superiority is an artifact of training and testing on labels generated by the same text-similarity and follow-graph signals the model uses.

Watch

Extended reading notes

Core claim

The central claim is that a three-module inductive architecture can outperform both content-only classifiers and a transductive method that enjoys an information advantage. IMMENSE represents each user's text as the summed Word2Vec embeddings of their posts, then trains two stacked autoencoders, one per class, whose reconstruction errors become the content signal. It feeds the same embeddings plus user labels into a supervised GraphSAGE model over the follow graph, and a second GraphSAGE model over a spatial-closeness graph built from z-normalized geodetic distances, producing safe/risky probabilities from each. A small MLP fuses these seven features into a final prediction. In the best reported configuration (focal loss, embedding dimension 512), IMMENSE reaches F1 0.967 overall; against SAIRUS it improves F1 on all users by up to 9.41% under class weighting and 14.12% under focal loss, and F1 on risky users by up to 20.83% and 30.54%, respectively, while reducing inference time by up to 76.93%.

Load-bearing premise

The load-bearing premise is that the synthetic ground truth—users labeled risky when the cosine similarity between their summed word-vector posts and a risky-post vector is at least 0.88, with safe users relabeled risky when more than 10% of their follow relationships point to risky users—captures real-world risky status independently of the model's own features.

Editorial extensions

If this is right

  • A trained IMMENSE model can be applied directly to new users or new networks, removing the need to retrain node embeddings on every new graph; reported inference on about 7,500 users takes at most 12.4 seconds.
  • All three perspectives contribute: adding spatial closeness improves F1 in 5 of 6 evaluated configurations even though only 1,043 of 37,945 users have location data, and the full fusion gives the best risky-class performance.
  • The inductive setting is harder than the transductive one, yet IMMENSE reports higher F1 than the transductive SAIRUS while also cutting inference time by roughly three-quarters.
  • Class imbalance can be handled in two ways, class weighting or focal loss, and both yield balanced precision and recall on the risky class rather than collapsing to a safe-only classifier.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The synthetic ground truth uses the same textual embedding space and follow graph that the model consumes, so an independent-label evaluation, such as expert annotation or platform enforcement actions, would show whether the reported margins reflect genuine detection or self-consistency with the labeling rule.
  • Because the architecture treats each perspective as a plug-in encoder, swapping Word2Vec for contextual embeddings such as BERT-style encoders is a natural testable extension, at higher computational cost.
  • The spatial module's usefulness despite sparse geolocation hints that coarser proxies, such as timezone, IP-derived region, or interaction locality, could extend the same design to networks where precise coordinates are unavailable.
  • If the labels do track real-world risk, the model's low inference cost makes it usable as an online triage system that surfaces candidate users for human review rather than as an autonomous enforcement tool.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper proposes IMMENSE, an inductive multimodal classifier for identifying 'risky' users in social networks. It combines a content module (summed Word2Vec embeddings plus one-class autoencoders), a relational module (GraphSAGE on the follow graph), a spatial module (GraphSAGE on geodetic closeness), and an MLP fusion. The authors evaluate on a Twitter/X dataset and report that IMMENSE outperforms five state-of-the-art competitors, including the transductive SAIRUS, while also reducing inference time.

Significance. If the evaluation were valid, the paper would make a useful contribution: an inductive, reusable model for law-enforcement monitoring, with open-source code and a clear modular design. The three-perspective architecture is coherent, and the paper includes informative ablations and explanation analyses. However, the empirical claim rests on a ground-truth definition that is manufactured from the same signals the method consumes, so the reported superiority cannot be interpreted as evidence of real-world detection capability. The method itself is not invalidated, but the central evidence for it is.

major comments (3)
  1. [§4.1, Eq. (11)] The ground-truth labels are defined as cosine_similarity(v_u, v_R) ≥ δ, where v_u is the sum of Word2Vec embeddings of the user's posts and v_R is the sum over known risky posts. This is exactly the additive Word2Vec representation used by the content module in Section 3.1. The subsequent social relabeling rule (safe users with more than 10% of relationships to risky users are relabeled risky) uses the same directed follow graph that the relational module in Section 3.2 feeds to GraphSAGE. The evaluation therefore measures how well IMMENSE can re-learn the authors' own labeling rule, not how well it detects risky users under an independent definition. A trivial baseline that applies Eq. (11) plus the social relabeling rule to the test set would be expected to achieve near-perfect F1, and the paper never compares against such a baseline. This is a load-bearing threat to the central claim that IMMENSE outperforms five state-of-the-art competitors.
  2. [§4.1] The dataset is described as 'real-world' because it is built from Twitter/X posts and follow edges, but the risky/safe labels are entirely synthetic: they are produced by the threshold δ = 0.88, chosen after inspecting the similarity distribution, together with the 10% social relabeling rule. There is no manual annotation, no external benchmark, and no evaluation against an independent notion of riskiness. Consequently, the experimental results can only support the statement that IMMENSE fits this particular synthetic labeling procedure better than models using different representations; they do not support the abstract's claim that IMMENSE is superior for identifying malicious users in practice. To support the central claim, the authors would need an independent gold standard (e.g., expert annotation or a pre-existing dataset) and a baseline consisting of the labeling rule itself.
  3. [§4.3, Tables 4–6] The comparison with competitors is not sufficient to establish superiority even internally to the synthetic labels. MrSBC and Re3py classify all users as safe (risky recall 0) across every configuration, which suggests their hyperparameters or input representations may not have been properly adapted to this task; HGT is given a different text encoder (XLNet) than IMMENSE, and Rio-GNN is only compared at three embedding dimensions. Moreover, the reported improvements over SAIRUS are based on a single train/test split with no variance or significance testing, and the paper itself notes the comparison is 'inherently unfair in favor of SAIRUS' because of its transductive access. These issues do not invalidate the method by themselves, but they mean the claimed margins of 9.41% and 20.83% are not established as statistically meaningful.
minor comments (6)
  1. [§3.2, Eq. (4)] The weighting rule appears reversed: since label 0 denotes safe (Section 3.1), the condition 'if y_i = 0' should assign the safe weight, not the risky weight. Please check.
  2. [§3.3, Eq. (8)] The mean and standard deviation of pairwise distances sum over ordered pairs (a,b) with a≠b but divide by |N| rather than by the number of ordered pairs |N|(|N|−1). This makes the z-scores scale-dependent and should be corrected.
  3. [Table 4] The rows for MrSBC and Re3py are identical across all configurations; please clarify whether the reported configurations actually correspond to different inputs, and if so why the outputs do not change.
  4. [Tables 2–5] The superscript 'a' used to denote 'without the fusion module' is only explained in the body text and not in the table captions; please define it in each caption.
  5. [Conclusions] The text says 'four state-of-the-art inductive competitors' while the abstract and Section 4.2 list five competitors (including SAIRUS); please reconcile the count.
  6. [Fig. 3] The claim of 'very neat separation' is based on a visual inspection of t-SNE plots; a quantitative cluster-separation measure would be more informative.

Circularity Check

2 steps flagged · score 8.0 of 10

The ground-truth labels are generated from the same Word2Vec content embeddings and follow graph that IMMENSE consumes, so the reported superiority reduces to re-learning the labeling rule rather than validated detection.

  1. self definitional [Section 4.1 (Eq. 11) vs Section 3.1]
    "We then measured the similarity between 𝑣𝑅 and the embedding of each user using cosine similarity. If this similarity exceeded a specified threshold 𝛿, the user was labeled as risky in the ground truth, otherwise it was labeled as safe... More formally, for each user 𝑢, we assigned a label as: 𝑙𝑎𝑏𝑒𝑙(𝑢) = { 𝑟𝑖𝑠𝑘𝑦 if 𝑐𝑜𝑠𝑖𝑛𝑒_𝑠𝑖𝑚𝑖𝑙𝑎𝑟𝑖𝑡𝑦(𝑣𝑢, 𝑣𝑅) ≥ 𝛿 𝑠𝑎𝑓𝑒 otherwise (11)"

    The ground-truth label is a threshold on cosine similarity between v_u and v_R in a Word2Vec additive space. Section 3.1 defines the semantic module's input identically: 'the user’s embedding to be represented as the sum of the embeddings of the words in their posted contents.' Thus the target Y is a deterministic function of the exact content feature X the model consumes; reporting F1 on this label measures how well IMMENSE re-learns Eq. (11). No baseline applying Eq. (11) directly is compared, so the claimed superiority over competitors is not evidence against the labeling rule itself.

  2. fitted input called prediction [Section 4.1 (social relabeling) vs Section 3.2]
    "After the initial labeling based on the similarity with the 𝑣𝑅 semantic vector, we examined the network of relationships of users labeled as safe. Any of these users having more than 10% of their relationships with risky users were relabeled as risky."

    This relabeling injects the follow graph into the ground truth. Section 3.2 builds the relational module on exactly that 'network of relationships': 'we construct a graph where users are connected based on their relationships in the analyzed social network' and trains GraphSAGE with the user labels. Consequently the relational perspective is evaluated against labels that were themselves created from the same follow-graph signal, so the relational module's gains reflect reproducing the relabeling rule rather than an independent social-risk signal.

full rationale

The central empirical claim—'Experiments carried out on a real-world Twitter/X dataset showed the superiority of IMMENSE against five state of the art competitors'—is supported only by a synthetic labeling procedure. Section 4.1 defines risky status by Eq. (11), a cosine-similarity threshold in the additive Word2Vec embedding space, and then relabels safe users who have more than 10% of relationships with risky users. Section 3.1 uses exactly the same additive Word2Vec content embeddings as the semantic module's input, and Section 3.2 uses exactly the same follow graph, with those content embeddings as node features, to train the relational module. The labels are therefore a deterministic function of the very representations the method was designed to exploit. The paper never compares against the trivial baseline that directly applies Eq. (11) plus the 10% relabeling rule, which should obtain near-oracle F1; without that baseline, IMMENSE's reported risky F1 of 0.870–0.940 cannot support superiority as real-world detection capability. This is a circular evaluation rather than a self-citation issue: the self-citations to SAIRUS, Re3py, and Mr-SBC are not load-bearing in the derivation, and the method itself is a legitimate inductive architecture. The circularity lies in presenting predictions of a self-generated labeling rule as independent validation.

Assumptions & free parameters 5 free parameters · 5 assumptions · 1 invented entities

The central evaluation rests on the ground truth construction in Section 4.1, which introduces two label-defining parameters (delta and the relabeling threshold) and assumes that a pre-trained Word2Vec similarity to v_R is a valid proxy for riskiness. The content and relational modules rely on the same representations, creating a circular loop between features and labels.

free parameters (5)
  • Ground truth similarity threshold delta = 0.88
    Chosen by inspecting the distribution of cosine similarities to v_R to yield about 7% risky users (Section 4.1). This parameter defines the target labels and therefore directly conditions all reported performance.
  • Social relabeling threshold = 0.10 (10% of relationships with risky users)
    Safe users with more than 10% of relationships to risky users are relabeled risky (Section 4.1), injecting the follow graph into the ground truth; no sensitivity analysis is reported.
  • Closeness threshold for spatial edges = 0 (z-normalized distance < 0)
    Pairs with z(distance) >= 0 get closeness 0 and no edge (Section 3.3, Eq. 10). This binarizing choice affects the spatial graph and the resulting module output.
  • Focal loss hyperparameters alpha and gamma = alpha=0.25, gamma=2.0
    Taken from the original focal loss paper [47], not fitted here; listed because they affect the reported results but are standard choices.
  • Embedding dimensions k_c, k_r, k_s = 128, 256, 512
    Three settings chosen for comparability across perspectives, based on widespread use in related work; results vary with these values, but no hyperparameter search is performed on the test set.
assumptions (5)
  • standard math Word2Vec embeddings are additive, so summing word vectors gives a meaningful user representation
    Invoked in Section 3.1 for the content module and in Section 4.1 for ground truth construction; this property is approximate in practice.
  • domain assumption One-class autoencoders handle class imbalance better than binary classifiers
    Motivates the content module (Section 3.1) based on cited work [45]; not independently validated here.
  • domain assumption Geographic proximity among users is informative of shared risk behavior
    Used to build the spatial closeness graph (Section 3.3); the paper cites no direct empirical evidence for this premise beyond the evaluation itself.
  • domain assumption Users who follow many risky users are themselves risky
    Explicitly baked into the ground truth relabeling (Section 4.1); this is a substantive assumption about social contagion that the evaluation then treats as truth.
  • domain assumption The pre-trained Google Word2Vec similarity to v_R identifies risky users
    The ground truth label (Eq. 11) assumes that cosine similarity to a global risky vector in a pre-trained embedding space is a valid proxy for user riskiness, which is not established by any external benchmark or human annotation.
invented entities (1)
  • Global risky vector v_R
    purpose: Aggregate embedding of known risky posts used to define the ground truth labels in Eq. (11); it is the reference against which each user's content embedding is compared.
    v_R is a constructed aggregate with no falsifiable handle outside this paper; it defines the labeling rule and therefore the entire evaluation, but is not validated against any external benchmark.

how reviews work

0 comments
Cite this review

Pith. "Pith review of IMMENSE: Inductive Multi-perspective User Classification in Social Networks." pith.science (2026). https://pith.science/paper/BXB76RCK

@misc{pith2026260805259,
  author       = {Pith},
  title        = {Pith review of: IMMENSE: Inductive Multi-perspective User Classification in Social Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BXB76RCK}},
  note         = {Machine review of arXiv:2608.05259}
}
read the original abstract

Online social networks increasingly expose people to users who propagate discriminatory, hateful, and violent content. Young users, in particular, are vulnerable to exposure to such content, which can have harmful psychological and social repercussions. Given the massive scale of today's social networks, in terms of both published content and number of users, there is an urgent need for effective systems to aid Law Enforcement Agencies (LEAs) in identifying and addressing users that disseminate malicious content. In this work we introduce IMMENSE, a machine learning-based method for detecting malicious social network users. Our approach adopts a hybrid classification strategy that integrates three perspectives: the semantics of the users' published content, their social relationships and their spatial information. Such contextual perspectives potentially enhance classification performance beyond text-only analysis. Importantly, IMMENSE employs an inductive learning approach, enabling it to classify previously unseen users or entire new networks without the need for costly and time-consuming model retraining procedures. Experiments carried out on a real-world Twitter/X dataset showed the superiority of IMMENSE against five state of the art competitors, confirming the benefits of its hybrid approach for effective deployment in social network monitoring systems.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

54 extracted references · 45 canonical work pages

  1. [1]

    Pellicani, G

    A. Pellicani, G. Pio, D. Redavid, M. Ceci, SAIRUS: Spatially-aware identification of risky users in social networks, Inf. Fusion 92 (2023) 435–449, http://dx.doi. org/10.1016/J.INFFUS.2022.11.029

    work page doi:10.1016/j.inffus.2022.11.029 2023

  2. [2]

    Igawa, S

    R.A. Igawa, S. Barbon Jr., K.C.S. Paulo, G.S. Kido, R.C. Guido, M.L.P. Júnior, I.N. da Silva, Account classification in online social networks with lbca and wavelets, Inform. Sci. 332 (2016) 72–83

    work page 2016

  3. [3]

    W. Wu, M. Ghazali, S. Hazlin Huspi, A review of user profiling based on social networks, IEEE Access 12 (2024) 122642–122670, http://dx.doi.org/10.1109/ ACCESS.2024.3430987

    arXiv 2024

  4. [4]

    Balasubramaniam, S

    K. Balasubramaniam, S. Vidhya, N. Jayapandian, K. Ramya, M. Poongodi, M. Hamdi, G.B. Tunze, et al., Social network user profiling with multilayer semantic modeling using ego network, Int. J. Inf. Technol. Web Eng. (IJITWE) 17 (1) (2022) 1–14

    work page 2022

  5. [5]

    H. Ko, S. Lee, Y. Park, A. Choi, A survey of recommendation systems: recom- mendation models, techniques, and application fields, Electronics 11 (1) (2022) 141

    work page 2022

  6. [6]

    D. Li, H. Liu, Z. Zhang, K. Lin, S. Fang, Z. Li, N.N. Xiong, CARM: Confidence- aware recommender model via review representation learning and historical rating behavior in the online platforms, Neurocomputing 455 (2021) 283–296

    work page 2021

  7. [7]

    H. Liu, C. Zheng, D. Li, Z. Zhang, K. Lin, X. Shen, N.N. Xiong, J. Wang, Multi- perspective social recommendation method with graph representation learning, Neurocomputing 468 (2022) 469–481

    work page 2022

  8. [8]

    H. Liu, C. Zheng, D. Li, X. Shen, K. Lin, J. Wang, Z. Zhang, Z. Zhang, N.N. Xiong, EDMF: Efficient deep matrix factorization with review feature learning for industrial recommender system, IEEE Trans. Ind. Inform. 18 (7) (2022) 4361–4371. Online Social Networks and Media 50 (2025) 100335 12 F. Benedetti et al

    work page 2022

Show all 54 references
  1. [9]

    Crime, J.R.I

    U.N.I. Crime, J.R.I. (UNICRI), Stop the virus of disinformation: The malicious use of social media by terrorist, violent extremist and criminal groups during the covid-19 pandemic - november 2020, 2020

    2020

  2. [10]

    Thompson, Radicalization and the use of social media, J

    R. Thompson, Radicalization and the use of social media, J. Strat. Secur. 4 (4) (2011) 167–190

    2011

  3. [11]

    Evkoski, A

    B. Evkoski, A. Pelicon, I. Mozetic, N. Ljubesic, P.K. Novak, Retweet communities reveal the main sources of hate speech, 2021, CoRR abs/2105.14898. arXiv: 2105.14898

    2021 arXiv

  4. [12]

    Alatawi, A.M

    H.S. Alatawi, A.M. Alhothali, K.M. Moria, Detecting white supremacist hate speech using domain specific word embedding with deep learning and BERT, IEEE Access 9 (2021) 106363–106374, http://dx.doi.org/10.1109/ACCESS.2021. 3100435

    2021 doi

  5. [13]

    M. Ji, Y. Sun, M. Danilevsky, J. Han, J. Gao, Graph regularized transductive classification on heterogeneous information networks, in: J.L. Balcázar, F. Bonchi, A. Gionis, M. Sebag (Eds.), Machine Learning and Knowledge Discovery in Databases, European Conference, ECML PKDD 20...

    2010 doi

  6. [14]

    Berkani, S

    L. Berkani, S. Belkacem, M. Ouafi, A. Guessoum, Recommendation of users in social networks: A semantic and social based classification approach, Expert. Syst. J. Knowl. Eng. 38 (2) (2021) http://dx.doi.org/10.1111/EXSY.12634

    2021 doi

  7. [15]

    Desrosiers, G

    C. Desrosiers, G. Karypis, Within-network classification using local structure similarity, in: W. Buntine, M. Grobelnik, D. Mladenić, J. Shawe-Taylor (Eds.), Machine Learning and Knowledge Discovery in Databases, Springer Berlin Heidelberg, Berlin, Heidelberg, 2009, pp. 260–275

    2009

  8. [16]

    Chapelle, V

    O. Chapelle, V. Vapnik, J. Weston, Transductive inference for estimating values of functions, in: S.A. Solla, T.K. Leen, K. Müller (Eds.), Advances in Neural Information Processing Systems 12, [NIPS Conference, Denver, Colorado, USA, November 29 - December 4 1999], The MIT Pre...

    1999

  9. [17]

    Z. Xue, Z. Zhang, H. Liu, S. Yang, S. Han, Learning knowledge graph embedding with multi-granularity relational augmentation network, Expert Syst. Appl. 233 (2023) 120953

    2023

  10. [18]

    Zhang, Z

    Z. Zhang, Z. Li, H. Liu, N.N. Xiong, Multi-scale dynamic convolutional network for knowledge graph embedding, IEEE Trans. Knowl. Data Eng. 34 (5) (2022) 2335–2347

    2022

  11. [19]

    Z. Li, H. Liu, Z. Zhang, T. Liu, N.N. Xiong, Learning knowledge graph embedding with heterogeneous relation attention networks, IEEE Trans. Neural Netw. Learn. Syst. 33 (8) (2022) 3961–3973

    2022

  12. [20]

    Z. Xue, Z. Zhang, H. Liu, Z. Li, S. Han, E. Zhang, Mhrn: A multi-perspective hierarchical relation network for knowledge graph embedding, Knowl.-Based Syst. 313 (2025) 113040

    2025

  13. [21]

    Perozzi, R

    B. Perozzi, R. Al-Rfou, S. Skiena, Deepwalk: Online learning of social represen- tations, in: Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2014, pp. 701–710

    2014

  14. [22]

    Grover, J

    A. Grover, J. Leskovec, Node2vec: Scalable feature learning for networks, in: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016, pp. 855–864

    2016

  15. [23]

    Hamilton, Z

    W.L. Hamilton, Z. Ying, J. Leskovec, Inductive representation learning on large graphs, in: I. Guyon, U. von Luxburg, S. Bengio, H.M. Wallach, R. Fergus, S.V.N. Vishwanathan, R. Garnett (Eds.), Advances in Neural Information Processing Systems 30: Annual Conference on Neural I...

    2017

  16. [24]

    Z. Hu, Y. Dong, K. Wang, Y. Sun, Heterogeneous graph transformer, in: Proceedings of the Web Conference 2020, 2020, pp. 2704–2710

    2020

  17. [25]

    J. Xiao, Q. Dai, X. Xie, J. Lam, K. Kwok, Adversarially regularized graph attention networks for inductive learning on partially labeled graphs, Knowl.-Based Syst. 268 (2023) 110456, http://dx.doi.org/10.1016/J.KNOSYS.2023.110456

    2023

  18. [26]

    H. Peng, R. Zhang, Y. Dou, R. Yang, J. Zhang, P.S. Yu, Reinforced neighborhood selection guided multi-relational graph neural networks, ACM Trans. Inf. Syst. (TOIS) 40 (4) (2021) 1–46

    2021

  19. [27]

    Velickovic, G

    P. Velickovic, G. Cucurull, A. Casanova, A. Romero, P. Liò, Y. Bengio, Graph attention networks, in: 6th International Conference on Learning Representations, ICLR 2018, Vancouver, BC, Canada, April 30 - May 3 2018, Conference Track Proceedings, OpenReview.net, 2018

    2018

  20. [28]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A.N. Gomez, L. Kaiser, I. Polosukhin, Attention is all you need, in: I. Guyon, U. von Luxburg, S. Bengio, H.M. Wallach, R. Fergus, S.V.N. Vishwanathan, R. Garnett (Eds.), Advances in Neural Information Processing Syste...

    2017

  21. [29]

    Baek, D.B

    J. Baek, D.B. Lee, S.J. Hwang, Learning to extrapolate knowledge: Transductive few-shot out-of-graph link prediction, in: H. Larochelle, M. Ranzato, R. Hadsell, M. Balcan, H. Lin (Eds.), Advances in Neural Information Processing Systems 33: Annual Conference on Neural Informat...

    2020

  22. [30]

    Hu, A two-step method for classifying political partisanship using deep learning models, Soc

    L. Hu, A two-step method for classifying political partisanship using deep learning models, Soc. Sci. Comput. Rev. 42 (4) (2024) 961–976

    2024

  23. [31]

    Z. Wu, S. Pan, F. Chen, G. Long, C. Zhang, P.S. Yu, A comprehensive survey on graph neural networks, IEEE Trans. Neural Netw. Learn. Syst. 32 (1) (2021) 4–24, http://dx.doi.org/10.1109/TNNLS.2020.2978386

    2021

  24. [32]

    Aslan, C

    H.I. Aslan, C. Choi, H. Ko, Classification of vertices on social networks by mul- tiple approaches, 2023, http://dx.doi.org/10.48550/ARXIV.2301.11288, CoRR abs/2301.11288. arXiv:2301.11288

    2023 doi arXiv

  25. [33]

    L. He, C. Lu, J. Ma, J. Cao, L. Shen, P.S. Yu, Joint community and structural hole spanner detection via harmonic modularity, in: B. Krishnapuram, M. Shah, A.J. Smola, C.C. Aggarwal, D. Shen, R. Rastogi (Eds.), Proceedings of the 22nd ACM SIGKDD International Conference on Kno...

    2016

  26. [34]

    J. Zhou, G. Cui, S. Hu, Z. Zhang, C. Yang, Z. Liu, L. Wang, C. Li, M. Sun, Graph neural networks: A review of methods and applications, AI Open 1 (2020) 57–81, http://dx.doi.org/10.1016/J.AIOPEN.2021.01.001

    2020 doi

  27. [35]

    Souza, R.M

    R.C. Souza, R.M. Assunção, D.M. Oliveira, D.B. Neill, W. Meira Jr., Where did i get dengue? detecting spatial clusters of infection risk with social network data, Spat. Spatio-Temporal Epidemiology 29 (2019) 163–175

    2019

  28. [36]

    Q. Gong, Y. Liu, J. Zhang, Y. Chen, Q. Li, Y. Xiao, X. Wang, P. Hui, Detecting malicious accounts in online developer communities using deep learning, IEEE Trans. Knowl. Data Eng. 35 (10) (2023) 10633–10649, http://dx.doi.org/10. 1109/TKDE.2023.3237838

    2023

  29. [37]

    J. Wang, X. He, Q. Gong, Y. Chen, T. Wang, X. Wang, Deep learning based mali- cious account detection in the momo social network, in: 2018 27th International Conference on Computer Communication and Networks, ICCCN, IEEE, 2018, pp. 1–2

    2018

  30. [38]

    Petkovic, M

    M. Petkovic, M. Ceci, G. Pio, B. Skrlj, K. Kersting, S. Dzeroski, Relational tree ensembles and feature rankings, Knowl.-Based Syst. 251 (2022) 109254, http://dx.doi.org/10.1016/J.KNOSYS.2022.109254

    2022

  31. [39]

    M. Ceci, A. Appice, D. Malerba, Mr-sbc: A multi-relational naïve bayes classifier, in: N. Lavrac, D. Gamberger, H. Blockeel, L. Todorovski (Eds.), Knowledge Discovery in Databases: PKDD 2003 7th European Conference on Principles and Practice of Knowledge Discovery in Databases...

    2003

  32. [40]

    Mikolov, K

    T. Mikolov, K. Chen, G. Corrado, J. Dean, Efficient estimation of word representa- tions in vector space, in: Y. Bengio, Y. LeCun (Eds.), 1st International Conference on Learning Representations, ICLR 2013, Scottsdale, Arizona, USA, May (2013) 2-4, Workshop Track Proceedings, 2013

    2013

  33. [41]

    Mikolov, I

    T. Mikolov, I. Sutskever, K. Chen, G.S. Corrado, J. Dean, Distributed representa- tions of words and phrases and their compositionality, Adv. Neural Inf. Process. Syst. 26 (2013)

    2013

  34. [42]

    Martino, G

    G.D. Martino, G. Pio, M. Ceci, PRILJ: An efficient two-step method based on embedding and clustering for the identification of regularities in legal case judgments, Artif. Intell. Law 30 (3) (2022) 359–390, http://dx.doi.org/10.1007/ S10506-021-09297-1

    2022

  35. [43]

    De Martino, G

    G. De Martino, G. Pio, M. Ceci, Multi-view overlapping clustering for the identification of the subject matter of legal judgments, Inform. Sci. 638 (2023) 118956

    2023

  36. [44]

    Rumelhart, J.L

    D.E. Rumelhart, J.L. McClelland, P.R. Group, et al., Parallel Distributed Processing, Foundations 1, 1988

    1988

  37. [45]

    Bellinger, S

    C. Bellinger, S. Sharma, N. Japkowicz, One-class versus binary classification: Which and when? in: 2012 11th International Conference on Machine Learning and Applications, vol. 2, 2012, pp. 102–106, http://dx.doi.org/10.1109/ICMLA. 2012.212

    2012 doi

  38. [46]

    Krawczyk, Learning from imbalanced data: open challenges and future directions, Prog

    B. Krawczyk, Learning from imbalanced data: open challenges and future directions, Prog. Artif. Intell. 5 (4) (2016) 221–232

    2016

  39. [47]

    T.-Y. Lin, P. Goyal, R. Girshick, K. He, P. Dollár, Focal loss for dense object detection, in: Proceedings of the IEEE International Conference on Computer Vision, 2017, pp. 2980–2988

    2017

  40. [48]

    Nazir, M

    S. Nazir, M. Asif, S.A. Sahi, S. Ahmad, Y.Y. Ghadi, M.H. Aziz, Toward the development of large-scale word embedding for low-resourced language, IEEE Access 10 (2022) 54091–54097

    2022

  41. [49]

    McVicar, B

    M. McVicar, B. Di Giorgi, B. Dundar, M. Mauch, Lyric document embeddings for music tagging, 2021, arXiv preprint arXiv:2112.11436

    2021 arXiv

  42. [50]

    L. Dong, N. Yang, W. Wang, F. Wei, X. Liu, Y. Wang, J. Gao, M. Zhou, H.-W. Hon, Unified language model pre-training for natural language understanding and generation, Adv. Neural Inf. Process. Syst. 32 (2019)

    2019

  43. [51]

    Z. Chen, H. Mao, H. Li, W. Jin, H. Wen, X. Wei, S. Wang, D. Yin, W. Fan, H. Liu, et al., Exploring the potential of large language models (llms) in learning on graphs, ACM SIGKDD Explor. Newsl. 25 (2) (2024) 42–61

    2024

  44. [52]

    Z. Yang, Z. Dai, Y. Yang, J. Carbonell, R.R. Salakhutdinov, Q.V. Le, Xlnet: Generalized autoregressive pretraining for language understanding, Adv. Neural Inf. Process. Syst. 32 (2019)

    2019

  45. [53]

    Maaten, G

    L.v.d. Maaten, G. Hinton, Visualizing data using t-sne, J. Mach. Learn. Res. 9 (2008) 2579–2605

    2008

  46. [54]

    Ribeiro, S

    M.T. Ribeiro, S. Singh, C. Guestrin, Why should i trust you? explaining the pre- dictions of any classifier, in: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016, pp. 1135–1144. Online Social Networks and Media 50 (2025) ...

    2025

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.