Projects

Some selected projects:

• Harvesting Tables in the Wild
  • Status: September/2020 - Feb/2021
  • Code: LINK
  • About:
    • Pipeline to harvest high-quality tables from multiple sources such as web and common crawl.
    • Implemented highly-scalable distributed architecture using cloud-native tecnologies including multi-modal databases (ArangoDB) and message brokers (Kafka).

• Security-Elasticity dependency analytics for Industrial IoT Infrastructure
  • Status: June/2020 - Aug/2020
  • Prototype to be released soon ;)
  • About:
    • This framework is able to perform cross-layer cross -system dependency analytics inside an IIoT network.
    • Aids the IIoT developer in generating concrete dependency between two services inside the IIoT infrastructure.

• An implementation of Dynamic Adaptive DNN Surgery for Inference Acceleration on the Edge
  • Status: Complete
  • Code: LINK
  • About:
    • This is the implementation of DNN Surgery Light (DSL) algorithm originally published by Chuang Hu et. al. in IEEE INFOCOM 2019 [1].
    • It has been implemented in python3 using pytorch DNN library. This implementation uses VGG-16 as the underlying architecture

• C# implementation of Automatic TextSummarization and keyword extraction
  • Status: Complete
  • Code: LINK
  • About:
    • This is the C# implementation of Automatic TextSummarization and keyword extraction based on TextRank algorithm [2].
    • This project came out as an intiative to improve the open-source library for C# and is inspired by one of the popular TextRank implementations for Python.

References:

[1] C. Hu, W. Bao, D. Wang and F. Liu, “Dynamic Adaptive DNN Surgery for Inference Acceleration on the Edge,” IEEE INFOCOM 2019 - IEEE Conference on Computer Communications, Paris, France, 2019, pp. 1423-1431, doi: 10.1109/INFOCOM.2019.8737614.

[2] R Mihalcea, P Tarau; Proceedings of the 2004 conference on empirical methods in natural language processing; 2004; [URL]