Pith. sign in

REVIEW 4 major objections 5 minor 45 references

Power Plant Detection for Energy Estimation using GIS with Remote Sensing, CNN & Vision Transformers

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper claims that a GIS-CNN-ViT pipeline detects power plants in satellite imagery more accurately than pure deep learning and thereby supports energy estimation.

desk verdict The paper's headline claim is never tested: the only experiment is a standalone CNN with inconsistent metrics, so the hybrid GIS+CNN+ViT story remains a proposal without evidence. read the letter →

arxiv 2412.04986 v1 pith:JRBMYWLY submitted 2024-12-06 cs.LG cs.CV

Blessing Austin-Gabriel , Cristian Noriega Monsalve , Aparna S. Varde This is my paper · ORCID
classification cs.LGcs.CV
keywords powerplantdetectionenergyestimationremotesensingGeographicInformationSystemsconvolutionalneuralnetworksVisionTransformerssatelliteimageclassificationhybriddeeplearning
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 proposes a hybrid pipeline in which Geographic Information Systems (GIS), a convolutional neural network (CNN), and a Vision Transformer (ViT) work together to detect power plants in satellite imagery and support energy estimation. The intended division of labor is that GIS supplies spatial masks and interactive mapping, the CNN extracts localized features, and the ViT captures long-range dependencies, so the combination classifies plant types better than any single model. Results from the CNN stage of the pipeline report high precision and recall on a four-class imagery set, which the paper reads as evidence that the hybrid enhances classification. If the paper is right, automated detection could gradually replace manual surveying for monitoring energy infrastructure and planning renewable expansion.

What carries the argument

The carrier of the argument is the pipeline itself: GIS with remote sensing produces spatial masks and an interactive map layer; a CNN built from convolutional, pooling, and dense layers extracts local spatial hierarchies; and a ViT applies self-attention over entire images to capture long-range dependencies. The claimed mechanism is that the GIS masks bring geospatial context into the CNN-ViT stages, while the CNN and ViT compensate for each other's limits, local detail versus global structure, so the assembled model classifies power plants more accurately than its parts used separately.

What would settle it

The claim would be settled by running the same satellite-image classification task twice, once with the GIS spatial masks included and once without; if the mask-free version matches or beats the reported 93.69% precision and 98.19% recall, the hybrid's advantage disappears.

Watch

Extended reading notes

Core claim

The central claim is that coupling GIS remote sensing with a CNN-ViT architecture yields more accurate power plant detection than pure deep learning, and that this accuracy feeds energy estimation. The paper classifies satellite images into power plant types such as solar, hydro, natural gas, and bitumen using a pipeline of GIS-generated spatial masks, CNN feature extraction, and ViT global context. In the reported evaluation, the CNN component reaches 93.69% precision and 98.19% recall on its test set, which the paper reads as confirmation that the hybrid enhances classification. The conclusion extends the result to monitoring, maintenance assessment, and sustainable energy planning.

Load-bearing premise

The load-bearing premise is that the location masks produced by the GIS step actually make the CNN-ViT classifier more accurate; the paper states this but reports no experiment that isolates it.

Editorial extensions

If this is right

  • If the hybrid claim holds, power plant monitoring can shift from manual surveying to automated, scalable detection over high-resolution satellite imagery.
  • Classification outputs can be attached to individual plants, so detection accuracy translates directly into per-plant energy estimation and grid planning.
  • The GIS layer makes results easier to interpret spatially, which matters for stakeholder-facing decisions in energy management.
  • The same pipeline can absorb additional data types, such as infrared and multitemporal imagery, for operational-condition assessment.
  • The approach extends naturally to offshore wind monitoring, as the conclusion notes.

Reading between the lines

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

  • Reader inference: a decisive test not reported in the paper is an ablation that removes the GIS mask stage from the pipeline; until that comparison exists, the hybrid's advantage over the CNN alone is a proposal rather than a demonstrated effect.
  • Reader inference: if the GIS masks carry most of the gain, a simpler mask-conditioned CNN might reach similar accuracy at far lower computational cost than a full ViT pipeline.
  • Reader inference: the real-time aspiration points toward lightweight ViT variants or distillation, since full ViTs are parameter-heavy and the paper itself flags their computational load.
  • Reader inference: the same integration logic could be tested on other infrastructure detection tasks, such as hospital or industrial facility monitoring, provided geospatial mask layers exist.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes a hybrid pipeline that combines a Geographic Information System (GIS) with remote-sensing capabilities, a Convolutional Neural Network (CNN), and a Vision Transformer (ViT) to detect and classify power plants from USGS satellite imagery, framed as a step toward energy estimation. Section III describes the proposed architecture, in which GIS-provided spatial masks are fed into a CNN-ViT classifier; Section IV then reports an experiment on a four-class classification problem (bitumen, hydro, natural gas, solar). The reported experiment, however, trains only a small CNN (Table I lists only Conv2D, MaxPooling2D, Flatten, and Dense layers) and reaches a peak validation accuracy of 66.63% (Table II) while separately claiming test precision of 93.69% and recall of 98.19%. The conclusion asserts that the hybrid model outperformed pure deep-learning approaches, but no baseline or hybrid model is evaluated. The paper also discusses potential applications in energy estimation, real-time monitoring, and interpretability, but none of these applications is evaluated experimentally.

Significance. If the central claim were established—that GIS-provided spatial masks combined with a CNN-ViT classifier improve power plant detection and thereby support energy estimation—the result would be practically valuable for automated environmental monitoring and sustainable-energy planning. The manuscript as written, however, does not test that claim: no ViT branch, no GIS-mask input, and no CNN-ViT fusion mechanism appear in the experimental section; no baseline isolates the contribution of the proposed components; and no energy-estimation output is evaluated. The only executed experiment is a standard standalone CNN classifier whose headline metrics are internally difficult to reconcile. On the positive side, the paper uses real USGS data and states a clearly defined, falsifiable architecture proposal. Nonetheless, the submission as it stands reads as a position or proposal statement rather than a verification of the claimed hybrid system's performance.

major comments (4)
  1. [§IV (Table I) and Abstract] The central claim of the abstract—that the proposed GIS+CNN+ViT hybrid "is found to enhance classification"—is not supported by the experimental section. Table I lists only conv2d, max-pooling, flatten, and dense layers; there is no ViT branch, no spatial-mask channel from GIS, and no mechanism fusing CNN and ViT features anywhere in the model that was trained. The sentence in §IV stating that "several images were subjected to our model training using the hybrid approach" is therefore misleading: the experiment actually run is a standalone CNN, so the claimed hybrid advantage is untested.
  2. [§IV (Table II and metric bullets)] The reported metrics are internally inconsistent as presented. Validation accuracy peaks at 66.63% at epoch 8 (Table II), while the text reports test-set precision of 93.69% and recall of 98.19%. For a four-class problem, such high precision and recall are difficult to reconcile with a 66.63% validation accuracy unless the test set is very small, heavily imbalanced, or the metrics are computed differently; the paper provides no confusion matrix, no class distribution, no dataset size, and no description of the train/test split (including whether the reported precision and recall are macro- or micro-averaged). The headline numbers therefore cannot be verified as reported.
  3. [Second 'V. CONCLUSIONS' section] The conclusion states that "our experiments demonstrated that the hybrid model outperformed pure deep learning approaches with respect to both precision and recall," but no pure-deep-learning baseline, alternative architecture, or ablation study is presented anywhere in Section IV. Without a comparator, the claim of superiority over any other method is unsupported; at most, the paper shows that a small CNN reaches a certain accuracy on an unspecified dataset.
  4. [§III.C and §V] Two load-bearing premises of the proposal are asserted but never measured. First, §III.C states that GIS "provides spatial masks that can be fed into the rest of the pipeline for further analysis," but no experiment uses such masks or compares the pipeline with and without them, so there is no evidence that the GIS component contributes to accuracy. Second, the paper's stated purpose is power plant detection "to assist energy estimation," yet Section IV evaluates only four-class classification; no energy-production estimate, regression output, or estimation error is reported. The energy-estimation benefit remains an application prospect, not a result.
minor comments (5)
  1. [Section numbering] The manuscript contains three sections labeled V (Discussion and Applications, Conclusions, and Acknowledgements) plus a References section labeled VI; the numbering from Section V onward needs to be corrected.
  2. [§IV (figure references)] The text says "Figs. 5 and 6" show correctly classified plants in natural gas and nuclear categories, but the figures actually presented are Fig. 4 (natural gas) and Fig. 5 (nuclear), and there is no Fig. 6. Figure citations should be reconciled with the figure set.
  3. [§IV, Eq. (1)] Equation (1) is typeset incorrectly (for example, the LaTeX command '\frac' appears as literal text), and the explanation says "yi is the true label and yi` is the predicted probability," which does not match the formula's notation; a properly typeset cross-entropy expression with consistent notation is needed.
  4. [§II (reference accuracy)] Several references appear to describe different works than the text claims: [8] is listed as an oil-palm-tree detection paper but is cited as a GIS-remote-sensing framework for energy infrastructure mapping; [9] is an article on AI in healthcare but is cited as a CNN that detects thermal power plants with 94.7% accuracy; and [12] is a paper on graph convolutional networks but is cited as a prior hybrid CNN-ViT model for power plant detection. These citations should be replaced or corrected.
  5. [§IV (reproducibility)] The experimental section does not state the dataset size, the number of images per class, or the train/validation/test split; such details are required for the reported results to be reproducible.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: the proposed hybrid is asserted rather than derived, and the CNN-only experiment is mislabeled as hybrid evidence, but no reported equation or fitted parameter reduces to its own input.

full rationale

The paper's central claim is that a GIS+CNN+ViT hybrid improves power plant detection and energy estimation. Walking the claimed derivation chain, there is no circular step in the strict sense: no equation is defined in terms of its target, no fitted parameter is renamed as an independent prediction, and no uniqueness theorem or load-bearing self-citation forces the conclusion. The experimental section reports only a standalone CNN (Table I lists Conv2D, MaxPooling2D, Flatten, and Dense layers; no ViT branch and no GIS spatial-mask input), while the conclusion states that 'the hybrid model outperformed pure deep learning approaches.' That is an evidence mismatch and an unsupported generalization, not a circular reduction: the hybrid performance is never measured, so it cannot be equivalent to the CNN metrics by construction. The paper does contain many self-citations, but they are in related-work and application passages and do not carry the central premise. The mild score reflects the self-citation density and the gap between the asserted hybrid claim and the reported CNN-only experiment, not a derivation loop. Under the quoted-evidence standard, no specific circular step can be exhibited.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper introduces no fitted parameters or invented entities, but it rests on domain assumptions about data quality and on an untested claim about the value of the GIS component.

assumptions (3)
  • domain assumption Satellite imagery from USGS can be correctly labeled into four power plant classes (bitumen, hydro, natural gas, solar).
    The paper defines this classification task in Section IV but provides no label source, labeling procedure, or class balance information.
  • domain assumption The reported precision, recall, and validation accuracy are computed correctly on a representative test set.
    The metrics are stated without an evaluation protocol or error analysis, and their internal consistency is questionable given the accuracy of 66.63% versus precision of 93.69%.
  • ad hoc to paper GIS can provide spatial masks that meaningfully improve classification when fed into the CNN-ViT pipeline.
    This is asserted in Section III.C and Section V but never demonstrated in any experiment, making the central hybrid claim depend on an untested premise.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Power Plant Detection for Energy Estimation using GIS with Remote Sensing, CNN & Vision Transformers." pith.science (2026). https://pith.science/paper/JRBMYWLY

@misc{pith2026241204986,
  author       = {Pith},
  title        = {Pith review of: Power Plant Detection for Energy Estimation using GIS with Remote Sensing, CNN & Vision Transformers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JRBMYWLY}},
  note         = {Machine review of arXiv:2412.04986}
}
read the original abstract

In this research, we propose a hybrid model for power plant detection to assist energy estimation applications, by pipelining GIS (Geographical Information Systems) having Remote Sensing capabilities with CNN (Convolutional Neural Networks) and ViT (Vision Transformers). Our proposed approach enables real-time analysis with multiple data types on a common map via the GIS, entails feature-extraction abilities due to the CNN, and captures long-range dependencies through the ViT. This hybrid approach is found to enhance classification, thus helping in the monitoring and operational management of power plants; hence assisting energy estimation and sustainable energy planning in the future. It exemplifies adequate deployment of machine learning methods in conjunction with domain-specific approaches to enhance performance.

Figures

Figures reproduced from arXiv: 2412.04986 by the authors.

Figure 2
Figure 2. Hybrid Model with GIS, remote sensing, CNN and ViT for [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 35 canonical work pages

  1. [9]

    Multimodal Data Integration • Incorporate additional data sources: We can integrate other remote sensing data (infrared imagery, multitemporal data etc.) for enhanced operational condition assessment

    work page

  2. [12]

    Deep learning in remote sensing applications: A meta - analysis and review,

    L. Ma, Y. Liu, X. Zhang, Y. Ye, G. Yin, and B. A. Johnson, “Deep learning in remote sensing applications: A meta - analysis and review,” ISPRS Journal of Photogrammetry and Remote Sensing , vol. 152, pp. 166 –177, Jun. 2019, doi: 10.1016/J.ISPRSJPRS.2019.04.015

    doi:10.1016/j.isprsjprs.2019.04.015 2019

  3. [1]

    Deploy GIS with remote sensing technology to detect main objects in power plants

    work page

  4. [2]

    Develop a robust model utilizing CNNs and ViTs for effective feature detection in high-resolution images

    work page

  5. [3]

    Classify types of power plants (solar, nuclear, natural gas etc.) in USGS satellite imagery

    work page

  6. [4]

    assisting in energy estimation and more, to guide future decision-making

    Examine operational conditions in power plants for routine maintenance, damage assessment, etc. assisting in energy estimation and more, to guide future decision-making. This hybrid approach enhances accuracy of power plant detection as observed in our experiments. IV. EXPERIMENTAL EVALUATION We discuss the technical performance of the proposed hybrid mod...

    work page

  7. [5]

    MobileNets: Efficient Convolutional Neural Networks for M obile Vision Applications,

    A. G. Howard et al., “MobileNets: Efficient Convolutional Neural Networks for M obile Vision Applications,” Apr. 2017, Accessed: Oct. 03, 2024. [Online]. Available: http://arxiv.org/abs/1704.04861

    arXiv 2017

  8. [6]

    The first layer had 16 filters, followed by 32 and 64 filters in subsequent layers, allowing the model to capture progressively more complex features

    Convolutional Layers: The convolutional layers (3×33×3) with ReLU activations performed feature extraction by learning spatial hierarchies in the input images. The first layer had 16 filters, followed by 32 and 64 filters in subsequent layers, allowing the model to capture progressively more complex features

    work page

Show all 45 references
  1. [7]

    Pooling Layers: MaxPooling (2×22×2) was applied to reduce the spatial dimensions of the feature maps, making the model computationally efficient while retaining important features

  2. [8]

    This mapping is corresponding to the four power plant classes: BIT (bitumen), Hydro, Natural Gas, and Solar

    Fully Connect Layers: The flattened output of the convolutional layers was passed through a dense layer with 256 units and a ReLU activation function, followed by a final dense layer with 4 units and softmax activation. This mapping is corresponding to the four power plant cla...

  3. [10]

    • Automated maintenance scheduling: The predictive maintenance can harness sensor - based model outputs to schedule repairs

    Real-Time Monitoring and Maintenance • Real-time analysis: Image detection with remote sensing can propel real time power plant monitoring. • Automated maintenance scheduling: The predictive maintenance can harness sensor - based model outputs to schedule repairs. It can help ...

  4. [11]

    Explanation and Interpretability • Model explainability: GIS with remote sensing can help to add more explanation to model predictions based on CNN & ViTs to make results more easily interpretable to stakeholders in energy management. • Visualization Tools: Interactive visual ...

  5. [13]

    Deep learning for remote sensing data: A technical tutorial on the state of the art,

    L. Zhang, L. Zhang, and B. Du, “Deep learning for remote sensing data: A technical tutorial on the state of the art,” IEEE Geosci Remote Sens Mag , vol. 4, no. 2, pp. 22 –40, Jun. 2016, doi: 10.1109/MGRS.2016.2540798

    2016

  6. [14]

    Road network extraction,

    J. Wang, J. Song, M. Chen, and Z. Yang, “Road network extraction,” Int J Remote Sens, vol. 36, no. 12, pp. 3144– 3169, Jun. 2015, doi: 10.1080/01431161.2015.1054049

    2015

  7. [15]

    Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions,

    W. Wang et al., “Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions,” Proceedings of the IEEE International Conference on Computer Vision , pp. 548 –558, Feb. 2021, doi: 10.1109/ICCV48922.2021.00061

    2021

  8. [16]

    Comprehensive survey of deep learning in remote sensing: theories, tools, and challenges for the community,

    J. E. Ball, D. T. Anderson, and C. S. Chan, “Comprehensive survey of deep learning in remote sensing: theories, tools, and challenges for the community,” J Appl Remote Sens, vol. 11, no. 04, p. 1, Sep. 2017, doi: 10.1117/1.JRS.11.042609

    2017 doi

  9. [17]

    Optimizing Mobile Vision Transformers for Land Cover Classification,

    P. F. Rozario et al. , “Optimizing Mobile Vision Transformers for Land Cover Classification,” Applied Sciences 2024, Vol. 14, Page 5920, vol. 14, no. 13, p. 5920, Jul. 2024, doi: 10.3390/APP14135920

    2024 doi

  10. [18]

    Deep Learning in Remote Sensing: A Comprehensive Review and List of Resources,

    X. X. Zhu et al. , “Deep Learning in Remote Sensing: A Comprehensive Review and List of Resources,” IEEE Geosci Remote Sens Mag , vol. 5, no. 4, pp. 8 –36, Dec. 2017, doi: 10.1109/MGRS.2017.2762307

    2017

  11. [19]

    Deep Learning Based Oil Palm Tree Detection and Counting for High-Resolution Remote Sensing Images,

    W. Li, H. Fu, L. Yu, and A. Cracknell, “Deep Learning Based Oil Palm Tree Detection and Counting for High-Resolution Remote Sensing Images,” Remote Sensing 2017, Vol. 9, Page 22 , vol. 9, no. 1, p. 22, Dec. 2016, doi: 10.3390/RS9010022

    2017 doi

  12. [20]

    Artificial intelligence in healthcare: past, present and future,

    F. Jiang et al., “Artificial intelligence in healthcare: past, present and future,” Stroke Vasc Neurol, vol. 2, no. 4, pp. 230–243, Dec. 2017, doi: 10.1136/SVN-2017-000101

    2017 doi

  13. [21]

    AN IMAGE IS WORTH 16X16 WORDS: TRANSFORMERS FOR IMAGE RECO GNITION AT SCALE,

    A. Dosovitskiy et al. , “AN IMAGE IS WORTH 16X16 WORDS: TRANSFORMERS FOR IMAGE RECO GNITION AT SCALE,” ICLR 2021 - 9th International Conference on Learning Representations, 2021

    2021

  14. [22]

    Vision Transformers for Remote Sensing Image Classification,

    Y. Bazi, L. Bashmal, M. M. Al Rahhal, R. Al Dayil, and N. Al Ajlan, “Vision Transformers for Remote Sensing Image Classification,” Remote Sensing 2021, Vol. 13, Page 516, vol. 13, no. 3, p. 516, Feb. 2021, doi: 10.3390/RS13030516

    2021 doi

  15. [23]

    Graph Convolutional Networks for Hyperspectral Image Classification,

    D. Hong, L. Gao, J. Yao, B. Zhang, A. Plaza, and J. Chanussot, “Graph Convolutional Networks for Hyperspectral Image Classification,” IEEE Transactions on Geoscience and Remote Sensing, vol. 59, no. 7, pp. 5966– 5978, Jul. 2021, doi: 10.1109/TGRS.2020.3015157

    2021

  16. [24]

    A related context is that of textual data (some of which can be descriptive pertaining to images)

    are crucial aspects, as addressed by researchers in a CSPM (cloud security posture management tool). A related context is that of textual data (some of which can be descriptive pertaining to images). As time evolves, it is important to understand the evolution of terminology, ...

  17. [25]

    A deep learning framework for remote sensing image registration,

    S. Wang et al., “A deep learning framework for remote sensing image registration,” JPRS, vol. 145, pp. 148–164, Nov. 2018, doi: 10.1016/J.ISPRSJPRS.2017.12.012

    2018 doi

  18. [26]

    Multispectral drone data analysis on coastal 8 dunes

    B. Gonzalez-Moodie, S. Daiek, J. Lorenzo-Trueba, J. and A.S. Varde, “Multispectral drone data analysis on coastal 8 dunes”. IEEE International Conference on Big Data, 2021 (pp. 5903-5905). IEEE

    2021

  19. [27]

    Analyzing Land Use Change and Climate Data to Forecast Energy Demand for a Smart Environment

    A. Prasad, A. S. Varde, R. Gottimukkala, C. Alo, and P. Lal, “Analyzing Land Use Change and Climate Data to Forecast Energy Demand for a Smart Environment ”. I EEE International Renewable and Sustainable Energy Conference (IRSEC), 2021 (pp. 1-6)

    2021

  20. [28]

    Mining GIS data to predict urban sprawl

    A. Pampoore-Thampi, A.S. Varde, and D. Yu, “Mining GIS data to predict urban sprawl ”. ACM KDD 2014 Conference, Bloomberg Track, 2021 arXiv preprint arXiv:2103.11338,

    2014 arXiv

  21. [29]

    Supervised learning techniques in mobile device apps for Androids

    P. Basavaraju, and A. S. Varde, “Supervised learning techniques in mobile device apps for Androids ”. ACM SIGKDD Explorations, 2016, 18(2), pp.18-29

    2016

  22. [30]

    Using long short-term memory recurrent neural network in land cover classification on Landsat and Cropland data layer time series,

    Z. Sun, L. Di, and H. Fang, “Using long short-term memory recurrent neural network in land cover classification on Landsat and Cropland data layer time series,” Int J Remote Sens, vol. 40, no. 2, pp. 593 –614, Jan. 2019, doi: 10.1080/01431161.2018.1516313

    2019

  23. [31]

    The hidden Web, XML and the semantic Web: Scientific data management perspectives

    F. M. Suchanek, A.S. Varde, P. Nayak, and P. Senellart, “The hidden Web, XML and the semantic Web: Scientific data management perspectives ”. In Proceedings of the 14th International Conference on Extending Databas e Technology, ACM EDBT, 2011 (pp. 534-537)

    2011

  24. [32]

    I am guessing you can't recognize this: Generating adversarial images for object detection using spatial commonsense

    A. Garg, N. Tandon, and A. S. Varde, “I am guessing you can't recognize this: Generating adversarial images for object detection using spatial commonsense ”. Proceedings of the AAAI Conference on Art ificial Intelligence, 2020 (Vol. 34, No. 10, pp. 13789-13790)

    2020

  25. [33]

    CSK-SNIFFER: Commonsense Knowledge for Sniffing Object Detection Errors

    A. Garg, N. Tandon, and A. S. Varde, “CSK-SNIFFER: Commonsense Knowledge for Sniffing Object Detection Errors”. ACM EDBT/ICDT Workshops, 2022

    2022

  26. [34]

    Federated Learning Approach for Remote Sensing Scene Classification,

    B. Youssef et al. , “Federated Learning Approach for Remote Sensing Scene Classification,” Remote Sensing 2024, Vol. 16, Page 2194 , vol. 16, no. 12, p. 2194, Jun. 2024, doi: 10.3390/RS16122194

    2024 doi

  27. [35]

    Enhancing Cloud Security Posture for Ubi quitous Data Access with a Cybersecurity Framework Based Management Tool

    G. Coppola, A.S. Varde, and J. Shang, “Enhancing Cloud Security Posture for Ubi quitous Data Access with a Cybersecurity Framework Based Management Tool ”. In IEEE 14th Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON) , 2023, (pp. 0590-0594)

    2023

  28. [36]

    SITAC: discovering semantically identical temporally altering concepts in text archives

    A. Kaluarachchi, D. Roychoudhury, A. S. Varde, and G. Weikum, G., “SITAC: discovering semantically identical temporally altering concepts in text archives ”. Proceedings of the 14th International Conference on Extending Database Technology , ACM EDBT, 2011 (pp. 566-569)

    2011

  29. [37]

    Commonsense knowledge in machine intelligence

    N. Tandon, A. S. Varde, and G. de Melo, “Commonsense knowledge in machine intelligence ”. ACM SIGMOD Record, 2017, 46(4), pp.49-52

    2017

  30. [38]

    Information to wisdom: Commonsense knowledge extraction and compilation

    S. Razniewski, N. Tandon, and A.S. Varde, “Information to wisdom: Commonsense knowledge extraction and compilation”. Proceedings of the 14th ACM International Conference on Web Search and Data Mining , ACM WSDM, 2021 (pp. 1143-1146)

    2021

  31. [39]

    C. J. Conti, A. S. Varde, and W. Wang, “Robot action planning by commonsense knowledge in human -robot collaborative tasks. IEEE Internat ional IOT, Electronics and Mechatronics Conference (IEMTRONICS) , 2020 (pp. 1-7). IEEE

    2020

  32. [40]

    Robo- csk-organizer: Commonsense knowledge to organize detected objects for multipurpose robots

    R. Hidalgo, J. Parron, A.S. Varde, and W. Wang, “Robo- csk-organizer: Commonsense knowledge to organize detected objects for multipurpose robots ”, Springer IEMTRONICS conference, 2024 . arXiv preprint arXiv:2409.18385

    2024 arXiv

  33. [41]

    Enhancing autonomous vehicles with commonsense: Smart mobility in smart cities

    P. Persaud, A. S. Varde, and S. Robila, “Enhancing autonomous vehicles with commonsense: Smart mobility in smart cities ”. IEEE 29th international conference on tools with artifici al intelligence (ICTAI) , 2017 (pp. 1008- 1012)

    2017

  34. [42]

    Challenging research issues in data mining, databases and information retrieval

    A.S. Varde, “Challenging research issues in data mining, databases and information retrieval ”. ACM SIGKDD Explorations, 2009, 11(1), pp.49-52

    2009

  35. [43]

    Heterogeneous data and big data analytics. Automatic Control and Information Sciences

    L. Wang, “Heterogeneous data and big data analytics. Automatic Control and Information Sciences”, 2017, 3(1), pp.8-15

    2017

  36. [44]

    A survey of multi- modal knowledge graphs: Technologies and trends

    W. Liang, P.D. Meo, Y. Tang, and J. Zhu, “A survey of multi- modal knowledge graphs: Technologies and trends ”,

  37. [2024]

    ACM Computing Surveys, 56(11), pp.1-41

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.