Word N-Gram Based Classification for Data Leakage Prevention

Alneyadi, Sultan; Sithirasenan, Elankayer; Muthukkumarasamy, Vallipuram

doi:10.1109/trustcom.2013.71

Cited by 11 publications

(5 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The majority of the literature on sensitivity labelling is associated with Data Loss Prevention (DLP) systems [8][9][10][11][12], notably focusing on classifying sensitivity at the document level. Other research uses sensitivity classification for confidential information redaction on declassified documents [13][14][15], where classification is often performed at finer granularity that reaches the token level.…”

Section: Related Workmentioning

confidence: 99%

Using BiLSTM Networks for Context-Aware Deep Sensitivity Labelling on Conversational Data

Pogiatzis

Samakovitis

2020

Applied Sciences

View full text Add to dashboard Cite

Information privacy is a critical design feature for any exchange system, with privacy-preserving applications requiring, most of the time, the identification and labelling of sensitive information. However, privacy and the concept of “sensitive information” are extremely elusive terms, as they are heavily dependent upon the context they are conveyed in. To accommodate such specificity, we first introduce a taxonomy of four context classes to categorise relationships of terms with their textual surroundings by meaning, interaction, precedence, and preference. We then propose a predictive context-aware model based on a Bidirectional Long Short Term Memory network with Conditional Random Fields (BiLSTM + CRF) to identify and label sensitive information in conversational data (multi-class sensitivity labelling). We train our model on a synthetic annotated dataset of real-world conversational data categorised in 13 sensitivity classes that we derive from the P3P standard. We parameterise and run a series of experiments featuring word and character embeddings and introduce a set of auxiliary features to improve model performance. Our results demonstrate that the BiLSTM + CRF model architecture with BERT embeddings and WordShape features is the most effective (F1 score 96.73%). Evaluation of the model is conducted under both temporal and semantic contexts, achieving a 76.33% F1 score on unseen data and outperforms Google’s Data Loss Prevention (DLP) system on sensitivity labelling tasks.

show abstract

Section: Related Workmentioning

confidence: 99%

Using BiLSTM Networks for Context-Aware Deep Sensitivity Labelling on Conversational Data

Pogiatzis

Samakovitis

2020

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…Enterprise content aware DLP solution or DLP solution in general offers many approaches to monitor, protect sensitive data at their end points [19]. Generally these types of systems validate and provide authorization to applications whenever access or transfer of confidential data is required.…”

Section: Data Leakage Prevention Systemsmentioning

confidence: 99%

“…These measure are being necessary as usage of external storage devices are becoming unavoidable at the fast growing IT environment. These can be applied to even smartphones and all type of removable devices.The encrypted data can be decrypted only in the system containing the encryption software which was used earlier for encryption also the user must have the encryption key [19]. The data could not be viewed in any other system which does not possess the appropriate algorithm which was used for encryption.…”

Section: A Encryptionmentioning

confidence: 99%

See 1 more Smart Citation

Data Leakage Prevention System: A Systematic Report

Gowr. P*,

2019

IJRTE

View full text Add to dashboard Cite

Technology is growing exponentially and in the fast growing technology more fast processing devices are being introduced. A mobile phone could even handle high processing software and even a small SD card could store the complete data of an organization. In this fast growing it environment maintaining security of data is highly important as a small loss of data might create critical impact in the organization, hence preventing their sensitive data became a greatest challenge. Traditionally organisation implemented methods like framing policies in an organization, implementing Firewall, virtual private network at the endpoints, but still these method started lagging as the technology developed the methodology of data leakage and data theft also attained heights. Hence there was a need for some system which could prevent leakage of data and this could be done with the help of Data Leakage Prevention system (DLPs). The DLPs are capable of detect leakage in Data at any states namely Data in rest, Data in transit and Data in use which increases the need of the DLPs.This paper briefly explains about the carious DLPs available, their limitations, providing the information about the gaps in security providing enough awareness on new developers as well as researchers and professionals for developing a next generation DLP which is capable of preventing data loss.

show abstract

“…The authors apply statistical and signal processing techniques on traffic flow to generate signature and/or extract features for classification purposes. In [23], the authors investigate the use of N-grams for data classification purposes. Their method is based on using the N-grams frequency to classify documents in order to detect and prevent leakage of sensitive and confidential data.…”

Section: Related Workmentioning

confidence: 99%

A Novel Method to Detect Encrypted Data Exfiltration

He¹,

Zhang²,

Ma³

et al. 2014

2014 Second International Conference on Advanced Cloud and Big Data

View full text Add to dashboard Cite

Cloud computing's distributed architecture helps ensure service resilience and robustness. Meanwhile, the big data stored in the cloud are valuable and sensitive and they are becoming attractive targets of attackers. In real life, attackers can carry out attacks such as Advanced Persistent Threat (APT) to invade cloud infrastructure and steal cloud users' confidential data through encrypted transmission. Unfortunately, the most commonly used methods, e.g., Deep Packet Inspection (DPI), cannot detect encrypted data leakage efficiently. In this paper, we propose a novel method to detect encrypted data exfiltration for cloud. Generally speaking, the proposed method is composed of two steps. First, cloud providers analyze all outgoing network traffic and find out encrypted traffic. Second, cloud providers determine whether the encrypted traffic is launched by cloud users expectedly. If not, the encrypted traffic will be considered as data exfiltration. Specially, in the first step, DPI and entropy technology are used together to find out encrypted traffic efficiently and in the second step, we determine whether the encryption is expected or not through building cloud users' network behavior profile. We have carried out extensive experiments in real-world network environment and the experimental results validate the feasibility and effectiveness of our method.

show abstract

Word N-Gram Based Classification for Data Leakage Prevention

Cited by 11 publications

References 3 publications

Using BiLSTM Networks for Context-Aware Deep Sensitivity Labelling on Conversational Data

Using BiLSTM Networks for Context-Aware Deep Sensitivity Labelling on Conversational Data

Data Leakage Prevention System: A Systematic Report

A Novel Method to Detect Encrypted Data Exfiltration

Contact Info

Product

Resources

About