BIG DATA comes to research and journalism
WITH THE WORLD overwhelmed by the sheer volume of information, and people making decisions based on a diversity of sources of information, the media has become acutely aware of the importance of big data – a reality that is now challenging the news business.
Along has come the launch of BigKinds, a big-data analytics service based on a news archive of over 42 million articles from leading media companies. BigKinds provides a big-data set of news.
Manager Kim Su Ji told the 2017 KPF Journalism Conference Fellowship that the world is overloaded with all kinds of information and people are making decisions based on different sources of information. People’s thought processes today, and even the discussions within organisations, are mostly based on data, he said.
Examples of this data include high-volume sensor data, website content, and social networking information from channels such as Facebook and Twitter. This mass of information can be analysed computationally to reveal patterns, trends, and associations, especially relating to human behaviour and interactions.
“However, with so much information available, how can we have access to data and use the data effectively?” asked Kim. “The question is what makes people turn to what is called big data, so that it becomes a trend society.”
Launched in April 2016, BigKinds has attracted the attention of media companies that used their service to produce more than 100 analytical news articles. In academia, about 40 journal articles and research reports have also drawn on the service.
Korean Press Foundation’s Choi Yoon, who is in charge of managing BigKinds’, said describes the company as a “news big-data analysis service” combining big-data analysis technology with the largest article database including comprehensive daily newspapers, business publications, regional daily newspapers, and broadcasters.
BigKinds reprocesses dozens of years of accumulated news content and provides a three-dimensional view of today’s news, words, and networks, he said.
It is a free service that anyone can sign up for and allows for analytics data downloads, public data fusion, and personalisation services.
Users can download various news meta-data such as names of persons, institutions and places, noun keywords, core keywords and classification information.
If users are not familiar with data analysis, they can use BigKinds’s visualisation tool to conduct a simple analysis.
The service’s intent was that users could see the main issues of the news gathered daily, including major people and top keywords, and do so at a glance. For example, “today’s issue” function automatically picks up the day’s issue within a selected area from daily collected news, while “today’s figures”, can analyse daily news sources and quotations gathered from the daily news and highlight the sources that have been cited by journalists. Meanwhile, “verbal” is a service that shows the comments made by relevant people about specific issues, said Kim.
Additional features include “field keyword”, which analyses people, institutions, and places by keyword in daily news, and “network” which connects “today’s issue” with the network of people, institutions, and locations related to the item.
“Keyword trend is a service that allows users to view the number of articles related to their search terms in a graph,” said Kim.
Those who want to dig for connections, or visually represent an issue, have several options that could help a researcher, journalist, or graphic artist improve their communication game.
For example, a researcher could turn to the “analysis of related words” feature that can check for keywords related to a search word through a word cloud. Similarly a journalist could lean on the “network analysis”, which coughs up the people, place and organisational network related to a search term. And a graphic artist could up their game through “public data fusion,” which can present data in an easily understood visual format, including news, stock quotes and demographics.