Entity Matching in Online Social Networks

2017

With the large number of users connected to social networks, screenname duplication is a rising problem, which leads to interference when trying to recognize users. A number of algorithms have been proposed to distinguish user profiles on one or multiple social networks. The main task in this context is to have robust features. According to the state-of-the-art approaches, features can be: content and behavioural based features, that compare content similarity between posts or behaviour similarity (timestamps between posts (behavioural), or overlapping between content (content) for example). Attribute-based features that compare profiles attributes, such as gender, age, location or image. In this paper, we tackle this problem and propose SocialMatching++ a novel approach that leverages: (1) user life events such as graduation, marriage or new job, which used to enhance the behavioural approaches (2) profile biographies, which consist in small paragraphs that users write to comprise ...

Lecture Notes in Computer Science, 2014

Piecing together social signals from people in different online social networks is key for downstream analytics. However, users may have different usernames in different social networks, making the linkage task difficult. To enable this, we explore a probabilistic approach that uses a domain-specific prior knowledge to address this problem of online social network user profile linkage. At scale, linkage approaches that are based on a naïve pairwise comparisons that have quadratic complexity become prohibitively expensive. Our proposed threshold-based canopying framework-named OPL-reduces this pairwise comparisons, and guarantees a upper bound theoretic linear complexity with respect to the dataset size. We evaluate our approaches on real-world, large-scale datasets obtained from Twitter and Linkedin. Our probabilistic classifier integrating prior knowledge into Naïve Bayes performs at over 85% F1-measure for pairwise linkage, comparable to state-of-the-art approaches.

2021

Identity matching is the process of mapping profile information from disparate data sources to one single entity; this is a crucial task for many businesses and governments. Introduction of Web 2.0 and the ever increasing number of social media platforms has led to an explosive amount of user participation and collaboration on web. An ordinary user has more than one social media profile, each of which has a unique set of properties and features. This thesis proposes a framework that uses syntactic and semantic based identity matching approaches among Facebook, Linkedin and Twitter user profiles. The framework accomplishes this task by collecting available profile data and performing analysis and comparison using a set of methodologies. These methods consist of weighted string matching techniques, Google Maps, YouTube and NLP web APIs. Extracted Profiles with a similarity score above a pre-computed threshold value are considered a match.

Journal of Interdisciplinary Mathematics, 2020

An online user creates accounts on numerous social platforms to fulfill his/her needs. The user shares interactive content through posts/tweets and communicates with friends and colleagues. In this paper, we inspect the problem of social accounts matching across social media services by proposing a two-step method. In the first step, the proposed method uses cross-link posts to get the seed users' accounts. After that, it utilizes the network features of the seed user and finds the association between the network relationships (follower and following) of the same user. By evaluating the experimental results, we found that the accuracy achieved by the proposed technique with the exact username is 98% and while the accuracy reduces to 45% when username across the sites is differed by one character.

2018

Today, over 2.62 billion people are active social media users accounting to one-third of the world's population. There exist hundreds of online social networking sites offering different services and functionality to their fellow user, however few among them are the most popular like Facebook, Twitter, Instagram and LinkedIn. Social networks are designed to address specific social needs, offering a distinct set of services and functionality. In order to enjoy the diverse range of services and to cover different facets of life, a user often registers on multiple social networks resulting in dissimilar identities of same users. The process of finding and linking those similar but disconnected identities of an individual scattered across social networks is termed as Identity Resolution or User identity Linkage. It has a significant impact on various problem domains such as recommendations, target marketing, user profiling, impersonator detection, etc. In this paper we propose a framework to the identity resolution problem. We also discuss various challenges of linking user's identities across online social networks.

Malaysian Journal of Computer Science, 2022

The establishment of connections among social network users using their profile information is an important task in social network analysis, which facilitates the development of various technological solutions such as stock market analysis, crime detection, tracking system of fraudulent events, etc. In this work, a proximity-based clustering method for networking LinkedIn profiles is presented. The proposed system computes proximity values between users using various attributes of user profiles. The proximity measures are computed by analyzing unstructured data of user profiles to connect users. The method addresses various issues such as comparison of familiar sentences, finding patterns, and sub-patterns among user profiles, assigning weights on attributes similarity, and computing total similarity which is associated with unstructured data. After computing proximity measures on various attributes of user profiles, the connecting edges between nodes are determined by employing artificial intelligence and a network graph is formed. The method is evaluated on a LinkedIn data set to form a connected graph. The strength of the proposed methodology lies in the formation of multi-layered network graphs, as it uses various attributes of the user profiles to connect them. The proposed methodology helps various applications like recommendation systems to form network graphs of selected attributes and perform the social network analysis. The method achieves an accuracy of 96%. However, the profiles containing abbreviations of important information are not matched and the system accuracy drops down in such cases.

Future Generation Computer Systems, 2018

h i g h l i g h t s • A User Identification Model solely based on User Generated Content across social sites is presented. • A solution to UGC-based user identification is proposed. • The importance of spatial, temporal and content similarity for user identification is identified. • The user identification experiments based on three ground truth datasets are conducted.

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering, 2009

Online Social Networks (OSN) are experiencing an explosive growth rate and are becoming an increasingly important part of people's lives. There is an increasing desire to aid online users in identifying potential friends, interesting groups, and compelling products to users. These networks have offered researchers almost total access to large corpora of data. An interesting goal in utilizing this data is to analyze user profiles and identify how similar subsets of users are. The current techniques for comparing users are limited as they require common terms to be shared by users. We present a simple and novel extension to a word-comparison algorithm [7], entitled Inter-Profile Similarity (IPS), which allows comparison of short text phrases even if they share no common terms. The output of IPS is simply a scalar value in [0, 1], with 1 denoting complete similarity and 0 the opposite. Therefore it is easy to understand and can provide a total ordering of users. We, first, evaluated the effectiveness of IPS with a user-study, and then applied it to datasets from Facebook and Orkut verifying and extending earlier results. We show that IPS yields both a larger range for the similarity value and obtains a higher value than intersection-based mechanisms. Both IPS and the output from the analysis of the two OSN should help to predict and classify social links, make recommendations, and annotate friends relations for social network analysis.

IEEE Access

In modern society, it is common for people to be active in many different online social networks at once. As new social network services arise every year, it remains a great challenge to integrate social data. Discovering multiple profiles of a single person across different social networks is a precondition for integration, but it is still challenging due to the inconsistency and disruption of the accessible information among social media networks (SMNs). Many studies have made efforts on user's profiles, users' contents, and network structure to address this issue, but the issue of how to consider all these information in a unified model and tackle them simultaneously still remains challenging. Considering that identical users tend to have partial similar friend relationship structures in different SMNs, especially friendship SMNs, we deepen the analysis of "friend" relationships (mutual following connections) in different SMNs, and propose PIFGM (Pairwise Identical Factor Graph Model), a novel factor graph model-based model, to address this problem by considering both user attributes and friend relationships across networks. We also present a distributed learning algorithm to handle large-scale social networks. We evaluate the proposed model on two different data collections: SNS and SR. Our experimental results validate the effectiveness and efficiency of the proposed model. The proposed PIFGM significantly outperforms several alternative methods by up to approximately 10%~20% in terms of F1 and precision on SNS and SR respectively. INDEX TERMS cross-platform, user profile matching, factor graph model, friend relationship, PIFGM method Lidong Wang was born in Wenzhou, Zhejiang, China in 1982. She received her Ph.D. degree from the College of Computer Science and Technology, Zhejiang University in 2013. She is currently an Associate Professor at Hangzhou Normal University. Her current research interests include image processing, machine learning and text mining.

Within this course of time, social media sites have gained a great attention. People rely on social media for different purposes. Not all the online social media will provide the same service. For this reason people tend to have multiple accounts on multiple social media sites. It is challenging and also interesting to identify the account that belongs to the same user in multiple social media sites. Many researchers have been conducted to match the user accounts in different social media sites. In this paper we have concentrated on different techniques used by the researchers to identify the similar accounts in multiple social media sites. Hence this paper will give an idea about the techniques that solves the problem of identifying user accounts across multiple online social media sites.