Analysing and Preventing Unintended Privacy Violations in Social Networks
Lead Research Organisation:
University of Edinburgh
Department Name: Sch of Informatics
Abstract
Privacy has many definitions, one of them is the right to conceal specific information from selected people. Online Social Networks (OSNs) have a large number of users that share billions of posts. The spread of OSNs brings privacy problems that are not addressed before. Users share personal data, photos, and videos about themselves as well as their friends.
These data can lead to unintended privacy violations such as friends sharing events about a user who did not want to share, unwanted people accessing information about a user by their friends' posts. Even the network of friends can help other people to guess if the user is a member of a group or not. Users have public data about themselves where other people can infer unintended information about them using the data. We are aiming to find and analyse such cases at first. After that, we will research ways to prevent these privacy violations. We will collect data from online social networks and process them for such cases. Birthday celebrations, congratulations that are given for nuptials and condolences for a family member's death are a few examples of events we are going to collect.
At the first stage, we are planning to use Twitter as our focus. We are going to use Twitter API to collect tweets with relevant keywords such as "happy birthday", "my condolences", and so on. After that, we are going to clean the text data. We are also planning to do sentiment analysis on the replies to classify whether they are positive or negative. To do the sentiment analysis, we will use machine learning algorithms. We are using python as our coding language and Tweepy to use the Twitter API efficiently. In the later stages, we also plan to organise a user survey to understand people's opinions about the types of privacy violations we found by our quantitative analysis.
These data can lead to unintended privacy violations such as friends sharing events about a user who did not want to share, unwanted people accessing information about a user by their friends' posts. Even the network of friends can help other people to guess if the user is a member of a group or not. Users have public data about themselves where other people can infer unintended information about them using the data. We are aiming to find and analyse such cases at first. After that, we will research ways to prevent these privacy violations. We will collect data from online social networks and process them for such cases. Birthday celebrations, congratulations that are given for nuptials and condolences for a family member's death are a few examples of events we are going to collect.
At the first stage, we are planning to use Twitter as our focus. We are going to use Twitter API to collect tweets with relevant keywords such as "happy birthday", "my condolences", and so on. After that, we are going to clean the text data. We are also planning to do sentiment analysis on the replies to classify whether they are positive or negative. To do the sentiment analysis, we will use machine learning algorithms. We are using python as our coding language and Tweepy to use the Twitter API efficiently. In the later stages, we also plan to organise a user survey to understand people's opinions about the types of privacy violations we found by our quantitative analysis.
Organisations
People |
ORCID iD |
Walid Magdy (Primary Supervisor) | |
Dilara Kekulluoglou (Student) |
Studentship Projects
Project Reference | Relationship | Related To | Start | End | Student Name |
---|---|---|---|---|---|
EP/N509644/1 | 01/10/2016 | 30/09/2021 | |||
2097898 | Studentship | EP/N509644/1 | 01/09/2018 | 31/03/2022 | Dilara Kekulluoglou |
EP/R513209/1 | 01/10/2018 | 30/09/2023 | |||
2097898 | Studentship | EP/R513209/1 | 01/09/2018 | 31/03/2022 | Dilara Kekulluoglou |