Personality-based Knowledge Extraction for Privacy-preserving Data Analysis

Vu, Xuan-Son; Jiang, Lili; Brändström, Anders; Elmroth, Erik

doi:10.1145/3148011.3154479

Cited by 8 publications

(8 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ahead of the above procedure, we apply NLP techniques [10] for raw data preprocessing (e.g., data normalization, standardization, non-utf8 characters removal etc.) and infer additional variables such as sentiment analysis and privacy concern based on the given variables.…”

Section: System Descriptionmentioning

confidence: 99%

“…It is clearly that privacyconcerns on social network is an important factor to protect people's privacy on personal data [9]. Since there was not any user defined privacy-concerns information in the given dataset, we apply an approach from [10] to infer people's privacy based on their FB status and personality. As shown in Figure 7, in Data Search, the Case-Study 3 Dashboard displays four different charts to characterize privacy-concerns of people across data sources and their demographic information.…”

Section: Case-study 3: Privacy-concern Analysismentioning

confidence: 99%

See 1 more Smart Citation

Graph-based Interactive Data Federation System for Heterogeneous Data Retrieval and Analytics

Ait-Mlouk

Elmroth

et al. 2019

The World Wide Web Conference

Self Cite

View full text Add to dashboard Cite

Given the increasing number of heterogeneous data stored in relational databases, file systems or cloud environment, it needs to be easily accessed and semantically connected for further data analytic. The potential of data federation is largely untapped, this paper presents an interactive data federation system (https://vimeo.com/ 319473546) by applying large-scale techniques including heterogeneous data federation, natural language processing, association rules and semantic web to perform data retrieval and analytics on social network data. The system first creates a Virtual Database (VDB) to virtually integrate data from multiple data sources. Next, a RDF generator is built to unify data, together with SPARQL queries, to support semantic data search over the processed text data by natural language processing (NLP). Association rule analysis is used to discover the patterns and recognize the most important co-occurrences of variables from multiple data sources. The system demonstrates how it facilitates interactive data analytic towards different application scenarios (e.g., sentiment analysis, privacyconcern analysis, community detection).

show abstract

Section: System Descriptionmentioning

confidence: 99%

Section: Case-study 3: Privacy-concern Analysismentioning

confidence: 99%

Graph-based Interactive Data Federation System for Heterogeneous Data Retrieval and Analytics

Ait-Mlouk

Elmroth

et al. 2019

The World Wide Web Conference

Self Cite

View full text Add to dashboard Cite

show abstract

“…Regarding personality prediction, the most influential Five Factor Model (FFM ) has become a standard model in psychology over the last 50 years [11]. Here we re-introduce a summary of Vu et al [9] regarding FFM. The five factors are defined as neuroticism, openness to experience, conscientiousness, agreeableness, and extraversion.…”

Section: The Five Factor Modelmentioning

confidence: 93%

“…-Evaluating the effectiveness of personality based privacy-guarantee through extensive experimental studies on a real UGC dataset. -Solving an imbalanced data distribution issue in privacy-concern detection raised by Vu et al [9] using an over-sampling approach.…”

Section: Introductionmentioning

confidence: 99%

Self-adaptive Privacy Concern Detection for User-Generated Content

Jiang

2023

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

To protect user privacy in data analysis, a state-of-the-art strategy is differential privacy in which scientific noise is injected into the real analysis output. The noise masks individual's sensitive information contained in the dataset. However, determining the amount of noise is a key challenge, since too much noise will destroy data utility while too little noise will increase privacy risk. Though previous research works have designed some mechanisms to protect data privacy in different scenarios, most of the existing studies assume uniform privacy concerns for all individuals. Consequently, putting an equal amount of noise to all individuals leads to insufficient privacy protection for some users, while over-protecting others. To address this issue, we propose a self-adaptive approach for privacy concern detection based on user personality. Our experimental studies demonstrate the effectiveness to address a suitable personalized privacy protection for cold-start users (i.e., without their privacy-concern information in training data).

show abstract

“…One of the challenges in this approach is to obtain user privacy concerns. Based on a previous work [27,26], we found that the privacybudget can be predicted using a strong correlation between user personality and their privacy concerns. Thus, we employed the model [26] to decide privacybudget of user-level data.…”

Section: Personalised Dp-embeddingmentioning

confidence: 95%

dpUGC: Learn Differentially Private Representation for User Generated Contents

Tran

Jiang

2019

Preprint

Self Cite

View full text Add to dashboard Cite

This paper firstly proposes a simple yet efficient generalized approach to apply differential privacy to text representation (i.e., word embedding). Based on it, we propose a user-level approach to learn personalized differentially private word embedding model on user generated contents (UGC). To our best knowledge, this is the first work of learning user-level differentially private word embedding model from text for sharing. The proposed approaches protect the privacy of the individual from re-identification, especially provide better trade-off of privacy and data utility on UGC data for sharing. The experimental results show that the trained embedding models are applicable for the classic text analysis tasks (e.g., regression). Moreover, the proposed approaches of learning differentially private embedding models are both framework-and dataindependent, which facilitates the deployment and sharing. The source code is available at https://github.com/sonvx/dpText.

show abstract

Personality-based Knowledge Extraction for Privacy-preserving Data Analysis

Cited by 8 publications

References 11 publications

Graph-based Interactive Data Federation System for Heterogeneous Data Retrieval and Analytics

Graph-based Interactive Data Federation System for Heterogeneous Data Retrieval and Analytics

Self-adaptive Privacy Concern Detection for User-Generated Content

dpUGC: Learn Differentially Private Representation for User Generated Contents

Contact Info

Product

Resources

About