From Symbols to Embeddings: A Tale of Two Representations in Computational Social Science

Abstract

Computational Social Science (CSS), aiming at utilizing computational methods to address social science problems, is a recent emerging and fast-developing field. The study of CSS is data-driven and significantly benefits from the availability of online user-generated contents and social networks, which contain rich text and network data for investigation. However, these large-scale and multi-modal data also present researchers with a great challenge: how to represent data effectively to mine the meanings we want in CSS?To explore the answer, we give a thorough review of data representations in CSS for both text and network.Specifically, we summarize existing representations into two schemes, namely symbol-based and embeddingbased representations, and introduce a series of typical methods for each scheme. Afterwards, we present the applications of the above representations based on the investigation of more than 400 research articles from 6 top venues involved with CSS. From the statistics of these applications, we unearth the strength of each kind of representations and discover the tendency that embedding-based representations are emerging and obtaining increasing attention over the last decade. Finally, we discuss several key challenges and open issues for future directions. This survey aims to provide a deeper understanding and more advisable applications of data representations for CSS researchers.

A New Unsupervised Technique to Analyze the Centroid and Frequency of Keyphrases from Academic Articles

“…This section carefully examines the experiment's findings. The proposed method splits the document length into eight (8) regions to analyze the keyphrase centroid and frequency (KCF). When a document's number of regions is increased by more than eight, the first region has a lower keyphrase frequency value as well as a lower centroid value than the eight-region.…”

“…The split() method is then applied to the key files to calculate the number of keyphrases based on the newline (\n) (also shown in Figure 2). The proposed technique then considers the length of the document is broken into eight (8) regions and employs the first appearance keyphrase to analyze the centroid and frequency of keyphrases.…”

“…"Despite this, they are employed in a broad range of Digital Information Processing applications including Information Retrieval, Digital Content Management, Recommender Systems [7], and Contextual Advertising" [2]. Among other things, it can be used for search engines, media searches, legal and geographic information retrieval, and digital libraries [2,8,9].…”

Machine learning meets the Journal of Public Budgeting and Finance: Topics and trends over 40 years

“…Topic modeling is one of the most popular text‐mining method to analyze publications and form themes from a large amount of textual data. STM is a widely used topic modeling strategy in computational social science (Bai et al, 2021; Chen et al, 2021; Kang et al, 2022). Similar to other common topic modeling methods such as latent Dirichlet allocation (LDA), STM assumes that each document is a mixture of multiple latent topics, with each topic having a different probability to appear in the document.…”

A Gaussian High-Dimensional Random Matrix-Based Method for Detecting Abnormal Student Behaviour in Chinese Language Classrooms

“…is method can effectively classify sea ice regions, but they only analyse sea ice data provided by the Canadian Ice Service, which is only applicable to sea ice image processing in specific situations and has a smaller application [10]. Chen et al proposed a kernel singular vector decomposition (KSVD) algorithm, which is based on a learning decomposition method to generate adaptive bases and extends the KSVD algorithm in nonlinear feature space [11]. An algorithm is an iterative approach that alternates between the sparse encoding of the kernel feature space based on a nonlinear dictionary and an update process for each atom in the dictionary, ending with a linear support vector machine classifier to obtain the classification accuracy [12].…”