Identifying future trends by podcast mining: an explorative approach for Web-based horizon scanning

Abstract

Purpose The aim of this paper is to gain knowledge in podcast mining as an additional source for Web-based horizon scanning (HS). The paper presents theoretical insights on the potential of podcast mining by exploring topics, which may be relevant in the future, and by reflecting the results against a background of HS approaches. The study provides a preliminary overview by presenting an exemplary list of podcast shows for further research. Design/methodology/approach The paper uses an exploratory quantitative content analysis, which was conducted on the basis of 30 topics deemed to be relevant in the future and which were identified in the field of applied science. Based on these topics, podcasts and episodes were identified which address future-oriented topics and were discussed in terms of range of content. Findings The findings indicate that future-oriented topics are addressed in podcasts. However, differences in dynamics and range of content of the podcasts concerned highlight the necessity of identifying a list of suitable podcasts according to the specific scanning focus and the dynamics of each future-oriented topics. Originality/value While a growing number of podcast studies have already noted the importance of podcasts as a key medium, for example, educational processes and media sciences, no detailed explanation of podcast mining as a tool for the purposes of HS has been published. The review therefore makes an original contribution to this field, highlighting areas where future research is needed.

Podcasting the future

“…Podcasts remain a potentially viable source of content to inform HS exercises, but at present remain a difficult medium to utilize. Whilst trials by Welz et al (2020) found that searches brought back relevant podcasts to their search requirements, in this case applied science, they did not manage to achieve the same level of results as other sources of data. As significant question remains, however.…”

“…Welz et al (2020) have looked at broadening the possible scope of HS by seeing whether podcasts can likewise be used as a data source. Podcasts have become an increasingly popular way for individuals and organizations to discuss, promote, and analyze literally anything.…”

“…A critical factor in the use of podcasts of HS is the ease or difficulty of mining them for key words and content. The work of Welz et al (2020) uses titles, key words, and summaries as the source for HS key word searches. This relies heavily on the accuracy and diligence of those who create the podcasts as well as those involved with the platforms and the process of listing and making podcasts available.…”

New perspectives for data‐supported foresight: The hybrid AI‐expert approach

“…In the second phase, it was the aim to further develop, continue, and standardize the process by integrating AI and Big Data and thereby making use of (partly‐) automated processes. The efforts included (i) further optimizing an already well‐established tool for scanning structured data from scientific databases, (ii) exploring unstructured data sources to retrieve new Spotlights using News‐Sites and Podcasts (Welz et al, 2021), and (iii) identifying market potentials for Spotlights by analyzing corporate press releases via ML algorithms.…”