Text Mining in Cybersecurity

Ignaczak, Luciano; Goldschmidt, Guilherme; Costa, Cristiano André da; Righi, Rodrigo da Rosa

doi:10.1145/3462477

Cited by 30 publications

(8 citation statements)

References 129 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The application of text mining in the field of cybersecurity is presented in this paper using a systematic literature review [6]. They found 2196 research using a methodical protocol, 83 of which were summarised and offered a taxonomy as a contribution to illustrate the many text mining-supported activities in the cybersecurity field [6].…”

Section: Literature Reviewmentioning

confidence: 99%

“…There is enormous potential to analyse the data with the aid of AI, ML, and NLP [6] due to the accumulation of a huge amount of cybercrime data, which may include complaint text or investigation description. Numerous applications in law enforcement, such as text summarisation, relationship extraction, crime prediction, criminal intelligence collecting etc., are now possible thanks to substantial research in the field of NLP [5].…”

Section: Literature Reviewmentioning

confidence: 99%

“…Due to BERT's capacity for comprehending text context, Google, and Wikipedia both acknowledge that every query now involves BERT. The prevention and detection of spam, intrusions, and online crimes also use text summary [5,6]. Text summarisation methods are applied to extract the summary of a potentially large collection of textual documents.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Retracted: Software solution for text summarisation using machine learning based Bidirectional Encoder Representations from Transformers algorithm

Abdulwahid

2023

IET Software

View full text Add to dashboard Cite

The data science has evolved over the past 2 decades, allowing the technical norms to be built in a way that can handle the new issues. While various technical issues develop, the requirement for text summary has always been there. Nearly 10 years ago, the foundation of automatic text summarisation was created, and since then, technical improvements and refinements have been made for large-scale big data handling, crime investigation, and cybersecurity, to name a few. There are several text summarising techniques, and these influence the outcomes as well. Another difference from the last 20 years is the requirement for time for text summarising. To pursue acquiring the findings, machine learning methods are applied to the core set of text phrases as a data set. Neural networks are now being used to improve text summarisation. But because there is more data available as short text to summarise, there will also be a big need for short text summarising. Utilising a quick summarising technique, accuracy, precision, and memory are improved. The focus of the challenge in this work is on brief text summarisation and boosting accuracy using a cutting-edge algorithm called Bidirectional Encoder Representations from Transformers (BERT). Bidirectional Encoder Representations from Transformers with transformer produced outstanding results for Short Text Summarisation. The model receives the input and performs a sequence-to-sequence analysis of the data down to the word level. The model's implementation is then contrasted with Word2Vec + RNN and Word2Vec + long short-term memory (LSTM), two earlier works. The proposed strategy, which seeks to increase the training data duration and accuracy for short text summarisation, produced best results utilising BERT + LSTM and BERT + Transformer. Using a confusion matrix to monitor and analyse the improved findings, it was shown that BERT + Transformer had an accuracy of 97%. The suggested model also performs better than existing models in terms of precision (46%), and recall (30%). K E Y W O R D Slearning (artificial intelligence), natural language processing, neural nets, software engineering, software maintenance, text analysisThis is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Section: Literature Reviewmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Retracted: Software solution for text summarisation using machine learning based Bidirectional Encoder Representations from Transformers algorithm

Abdulwahid

2023

IET Software

View full text Add to dashboard Cite

show abstract

“…Compared with the traditional arcuate baffle support for nuclear power, this support structure changes the shell-side fluid from the transversely swept tube bundle to the longitudinally swept tube bundle, which can effectively reduce the flow dead zone and reduce the vibration induced by the tube bundle. And due to the fluid jet generated in the hole of the support plate, the deposition of chemical substances, corrosive substances, etc., on the wall of the heat exchange tube can be reduced, so that the heat transfer and corrosion conditions in the area near the wall of the heat exchange tube are greatly improved; the probability of heat exchanger failure is reduced [12].…”

Section: Brief Introduction Of Three-leaf Orifice Plate Heatmentioning

confidence: 99%

Development of Intelligent CAD Technology for New Longitudinal Shell-Side Heat Exchange Equipment

Wu¹,

Zhu²

2022

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

In order to correct the errors of “periodic full-section calculation model” and “unit flow channel model,” the author proposes a numerical simulation method of a longitudinal flow shell-side heat exchanger based on CAD. The numerical simulation of the fluid flow and heat transfer characteristics of the three-blade orifice plate heat exchanger is carried out by using FLUENT software, and the “periodic full-section calculation model” and “unit flow channel model” are established, as well as the calculation results of comparative analysis. Experimental results show the following: With the increase of the inner diameter D of the heat exchanger shell, the calculation results of the two models are gradually reached. When D > 800 mm , the error of the calculation results of the two models has been reduced to about 10%; at this time, the “unit flow channel model” has practicability and applicability. When D ≤ 800 mm , the correction algorithm of the “unit flow channel model” is proposed, and the correction correlation formula of the pressure gradient and convective heat transfer coefficient is given. The revised results of the simulation can not only meet the engineering needs but also save computer resources and improve the calculation efficiency, which provides a theoretical basis for the development, improvement, and further industrial application of the “unit flow channel model” of the longitudinal shell-side heat exchanger.

show abstract

“…These features can be key words or phrases that often appear in documents. The modeling stage is done to build a classification or clustering model that can be used to predict or classify text data [15].…”

Section: Text Mining Techniquesmentioning

confidence: 99%

Exploring network security threats through text mining techniques: a comprehensive analysis

Wahyuningsih,

Sembiring,

Setiawan

et al. 2023

Comput Sci Inf Technol

View full text Add to dashboard Cite

In response to the escalating cybersecurity threats, this research focuses on leveraging text mining techniques to analyze network security data effectively. The study utilizes user-generated reports detailing attacks on server networks. Employing clustering algorithms, these reports are grouped based on threat levels. Additionally, a classification algorithm discerns whether network activities pose security risks. The research achieves a noteworthy 93% accuracy in text classification, showcasing the efficacy of these techniques. The novelty lies in classifying security threat report logs according to their threat levels. Prioritizing high-risk threats, this approach aids network management in strategic focus. By enabling swift identification and categorization of network security threats, this research equips organizations to take prompt, targeted actions, enhancing overall network security.

show abstract

Text Mining in Cybersecurity

Cited by 30 publications

References 129 publications

Retracted: Software solution for text summarisation using machine learning based Bidirectional Encoder Representations from Transformers algorithm

Retracted: Software solution for text summarisation using machine learning based Bidirectional Encoder Representations from Transformers algorithm

Development of Intelligent CAD Technology for New Longitudinal Shell-Side Heat Exchange Equipment

Exploring network security threats through text mining techniques: a comprehensive analysis

Contact Info

Product

Resources

About