Social network analysis is an important problem in data mining. A fundamental step for analyzing social networks is to encode network data into low-dimensional representations, i.e., network embeddings, so that the network topology structure and other attribute information can be effectively preserved. A recurrent criticism concerning the use of online social media data in political science research is the lack of demographic information about social media users.

In Deep Representation Learning for Social Network Analysis  authors conduct a comprehensive review of the current literature utilizing neural network models. They introduce the basic models for learning node representations in homogeneous networks, and also introduce some extensions of the base models, tackling more complex scenarios such as analyzing attributed networks, heterogeneous networks, and dynamic networks. Then authors introduce techniques for embedding subgraphs and also present the applications of network representation learning.

Example of node representation learning and subgraph representation learning
Example of node representation learning and subgraph representation learning

DeepInf: Social Influence Prediction with Deep Learning   is inspired by the recent success of deep neural networks in a wide range of computing applications, we design an end-to-end framework. DeepInf takes a user’s local network as the input to a graph neural network for learning her latent social representation. Authors design strategies to incorporate both network structures and user-specific features into convolutional neural and attention networks. Extensive experiments on different types of social and information networks, demonstrate that the proposed end-to-end model significantly outperforms traditional feature engineering based approaches, suggesting the effectiveness of representation learning for social applications.

To date, many of the studies on politics and social media have focused on Twitter, primarily because of the public nature of the tweets posted on the platform. The article  “Using deep-learning algorithms to derive basic characteristics of social media users: The Brexit campaign as a case study”achieves the goals of  testing the precision of the algorithm,  testing its validity,  inferring new evidence on digital mobilisation, and  tracing the path for future developments and application of the algorithm. The findings show that the algorithm is reliable and that it can be fruitfully used in political and social sciences both to confirm the validity of survey data and to obtain information from populations that are generally unavailable within traditional surveys.

Age and gender distribution detected with the CNN algorithm
Age and gender distribution detected with the CNN algorithm

In Prediction of Facebook Post Metrics using Machine Learning authors showed have predicted the impact of a post in social network Facebook. The compared prediction accuracy of three models – Support Vector Regression, Echo State Network and Adaptive Neuro-Fuzzy Inference System. The dataset contains 7 features known prior to post publication, and 3 output variables which are used for the post impact. The output variables are: comments, shares, and likes. The new propose techniques in this article (ESN and ANFIS) obtain better results than SVR. Although, ANFIS seems to performs better for predicting the amount of likes, the ESN model has a better accuracy in the other two cases.

Sentiment analysis on social media such as Twitter has become a very important and challenging task. Due to the characteristics of such data—tweet length, spelling errors, abbreviations, and special characters—the sentiment analysis task in such an environment requires a non-traditional approach. Moreover, social media sentiment analysis is a fundamental problem with many interesting applications. Many works had been performed on twitter sentiment analysis but there has not been much work done investigating the effects of location on twitter sentiment analysis. Most current social media sentiment classification methods judge the sentiment polarity primarily according to textual content and neglect other information on these platforms. In study “Tweet sentiment analysis using deep learning with nearby locations as features “authors concatenated text and location features as a feature vector for twitter sentiment analysis using a deep learning classification approach specifically Convolutional Neural Network (CNN). The achieved results show that using location as a feature alongside text has increased the sentiment analysis accuracy.

In “Twitter sentiment analysis with a deep neural network: An enhanced approach using user behavioral information” proposed a  Convolutional Neural Network (CNN) model that also incorporates user behavioral information within a given tweet. The system is evaluated on two datasets provided by the SemEval-2016 Workshop. The proposed model outperforms current  Naive Bayes and Support Vector Machines models, which shows that going beyond the content of a tweet is beneficial in sentiment classification.