A Study on the Research Trends in the Recent Human Resource Development: Topic Modeling Approach

Chad (Chungil Chae), Jamal Al Khadhuri, Hyunwoo Yang, Woongbae Park // Pennsylvania State University

chadchae@gmail.com

Introduction

Background

  • It is useful to study research trends in order to review the previous research foci and redefine what we study within Human Resource Development scholarly community.

  • it is not easy to define HRD within a few daily-used expressions, particularly in light of the multi-disciplinary nature of the field, but the study provides a baseline for where researchers have been and suggests where they may be going next.

Problem

  • Regrading the continuously changing research topics and environment, it is necessary to investigate recent research trends and examine the findings that are suggested by previous research

Motivation

  • A different approach to the same topic sometimes gave us a chance to have a different view on the same object.

  • Wang, Gilley and Sun (2012) stated that the scientometric theories and methods would be useful for disciplinary structure pattern of human resource development field.

Purpose of Study

  • The purpose of this research study is introduce the statistical methodology or topic modeling as an effective and valuable tool to analyze the trend and provide meaning of the topic published in the four AHRD journals in the last five years.

Research Questions

  • What is the meaningful trend in the topics that have been studied in AHRD four journals within the recent five years using topic modeling?

  • How have the extracted topics constitute the body of HRD Knowledge?

  • What trends do the topics reflect?

Literature Review

Literature

  • Ghosh et al (2014) examined this phenomenon by content analysis using AHRD’s four journals’ peer-reviewed articles that are published in 10 years’ time (2002-2011).

  • The Russ-Eft, Watkins, Marsick, Jacobs, & McLean (2014) Their study reflected on the perspectives of five influential HRD scholars about the past, the present, and what the next 25 years hold for HRD research.

  • Bierema and Callahan (2014) discussed that HRD have assumed what they describe as a “masculine” path because it focused on competitiveness and addressing the enhancement of organizational performance.

Methodology

Topic modeling is:

  • A useful tool in the quest to identify a hidden topic pattern in the literature. Topic modeling utilizes a set of algorithms to discover the hidden thematic structure in the large unstructured text data

With Topic Modeling:

  • Provides methods for automatically organizing, understanding, searching and summarizing large electronic archives without any prior annotation or labeling
  • Discover the hidden themes that pervade the collection
  • Annotate the documents according to those themes
  • Use annotations to organize, summarize, and search the text

Probablic Topic Modeling with Latent Dirichlet allocation

Developmental History of Topic Modeling

Probablic Topic Modeling with Latent Dirichlet allocation

A Generative model

  • Documents are mixture of topics, where a topic is a probability distribution over words
  • Genetics topic has words about genetics with high probability

Probablic Topic Modeling with Latent Dirichlet allocation

  • Each topic is a distribution over words
  • Each document is a mixture of corpus-wide topics
  • Each word is drawn from one of those topics

Data Collection

  • Abstracts from 746 articles
  • 4 AHRD sponsered journals (HRDQ, ADHR, HRDR, HRDI)
  • 6 years window (2010 - 2015)

Analystic Procedure

Findings

Clearning

  • 564 Documents
  • 657 terms
  • 26936 tokens

Topics

  • Total 30 topics identified
  • For the naming, two coders evaluated each topics
  • Multi coder rating, Kappa value were 0.829

Topics

Topics

Proportion of Topics

Topic Correlation (Network)

  • Positive value of correlation was reflected as ties

Conclusion

Limitation

  • The research trend by themes reflected only the recent five-year publications.
  • Generalizability: Articles from the four AHRD journals
  • For the more precise result, preparation stage should be improved
  • Evaulation / Validation of topic model

Conclusion

  • This study purposed to extract recent five years’ topics that the HRD scholarly community collectively generated, which shaped the body of HRD field of studies, and analyzes research topic trends and characteristics by text analysis known as topic modeling that is based on unsupervised machine learning technology.

  • Findings from this study identified dominant topics as themes that reflect socially constructed meaning based on co-occurrence of words

Question and Answer

Contact