From Manual Coding to Large-Scale Text Analysis: NLP in Computational Social Science

Introduction

Second Season of Computational Social Science Reading Club In the seventh session, we invite Dr. Jiaxin Pei from the University of Michigan, Ann Arbor, to introduce the basic process of Natural Language Processing (NLP) and focus on how to extend traditional manual coding to large-scale automated text analysis. Based on this, we will discuss the combination and application of NLP methods in fields such as communication and psychology using two research examples. The reading club will take place on August 14 (Sunday) from 9:00 AM to 11:00 AM, with a link for live registration provided later.

【Second Season of Computational Social Science Reading Club】 Led by Professor Luo Jiade from Tsinghua University, with the participation of several PhD students from Carnegie Mellon University, University of Michigan, Tsinghua University, and University of Pittsburgh, this reading club has been ongoing since June 18, 2022, lasting 10-12 weeks. This season will focus on discussing methods such as Graph, Embedding, NLP, Modeling, and Data Collection, and their integration with social science issues, specifically addressing topics like predictive and explanatory research, and studies related to the COVID-19 pandemic. Details and participation methods for the reading club can be found at the end of the text. Friends engaged in related research or interested in computational social science are welcome to register!

Content Overview

Relationship with the Computational Social Science Reading Club

Text analysis is a very important research method in social sciences. However, traditional manual coding methods severely limit the sample size that can be studied, making it difficult to apply to large-scale data analysis. In this session, we will introduce the entire process of text analysis methods based on NLP technology and use two articles as examples to demonstrate the application of NLP methods in computational social science research.

Main Knowledge Concepts Involved

Natural Language Processing, Text Analysis, Data Annotation

Introduction to the Session

In this session, we will introduce the basic process of NLP and focus on how to extend traditional manual coding to large-scale automated text analysis. Based on this, we will discuss the combination and application of NLP methods in fields such as communication and psychology using two research examples.

Session Outline

1. Basic process of Natural Language Processing

2. Data-driven manual coding

3. Research Case 1: Quantifying Language Intimacy

a. How to quantify language intimacy

b. How gender, social distance, and anonymity affect language intimacy

4. Research Case 2: Changes in Certainty in Scientific Communication

a. How to quantify the certainty of scientific discoveries

b. Has the certainty of scientific discoveries changed in scientific communication?

c. How do team size and journal impact factors affect the certainty of scientific discoveries?

Speaker Introduction

Jiaxin Pei, a PhD student in Information Science at the University of Michigan, Ann Arbor, focuses on computational social science and natural language processing, dedicated to building new NLP models to analyze large-scale human behavior, currently focusing on interpersonal communication and scientific communication, with relevant results published in conferences such as ACL, EMNLP, WWW.

Homepage link:

https://jiaxin-pei.github.io/

References

Literature discussed in this session

• Pei, Jiaxin, and David Jurgens. “Quantifying Intimacy in Language.” Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2020.
• Pei, Jiaxin, and David Jurgens. “Measuring Sentence-Level and Aspect-Level (Un) certainty in Science Communications.” Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing. 2021.

Related Literature

• Danescu-Niculescu-Mizil, Cristian, et al. “A computational approach to politeness with application to social factors.” Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2013.
• Golder, Scott A., and Michael W. Macy. “Diurnal and seasonal mood vary with work, sleep, and daylength across diverse cultures.” Science 333.6051 (2011): 1878-1881.

Live Stream Information

Live Stream Time:

August 14, 2022 (Sunday) from 9:00 AM to 11:00 AM

Participation Method:

• Free live streaming on Bilibili account of the Intelligence Club, scan the code to make a reservation:

Scan the code to reserve for this live stream

• If you need to watch the video replay, scan the code at the end to pay to participate in the second season of the Computational Social Science Reading Club, you can join Tencent Meeting, ask questions, join group chats, obtain video replays, and more study materials, becoming a seed user of the computational social science community, communicating with more than 450 frontline researchers in computational social science to promote the development of the community.

Recruitment for the Second Season of the Computational Social Science Reading Club

The second season of the Computational Social Science Reading Club is led by Professor Luo Jiade from Tsinghua University, with the participation of several PhD students from Carnegie Mellon University, University of Michigan, Tsinghua University, and University of Pittsburgh, ongoing since June 18, 2022, lasting 10-12 weeks. This season will focus on discussing methods such as Graph, Embedding, NLP, Modeling, and Data Collection, and their integration with social science issues, specifically addressing topics like predictive and explanatory research, and studies related to the COVID-19 pandemic. Details and participation methods for the reading club can be found at the end of the text. Friends engaged in related research or interested in computational social science are welcome to participate.

Click “Read the original text” to register for the live stream