Uncovering Opinions on Cybersecurity and Accessibility in Metaverse

Mikyung Shin (Assistant Professor, West Texas A&M University)

Gahangir Hossain (Associate Professor, University of North Texas)

Mehnaz Afrose (Graduate Research Assistant, West Texas A&M University)

  github.com/mshin77/GTM2023     mshin77.net

Agenda

• What is Metaverse?
  

• South Korea’s Innovation Toward the Metaverse
  

• Cybersecurity Issues in the Metaverse
  

• News Article Analysis

What is Metaverse?

• Meta (Beyond) + Verse (Universe)
• A network of 3D virtual worlds which are connected.
• Provide opportunities to connect and communicate with others.
• Feel like everyone is physically present with others while engaging in virtual experiences
• Digitally represent ourselves through an avatar (who we are, who we want to be)
• Establish interoperability (allow individuals to take things with them as they move around in the virtual spaces)

Source: AI for Good

South Korea’s Innovation Toward the Metaverse

Source: Google Trend

Cybersecurity Issues in the Metaverse

• Naver News: South Korean online platform operated by the Naver Corporation and launched in 1998 (news service, 2000-present, available open news search, 1990-present)

• Big KINDS: Integrated database for most major newspapers published in South Korea 1990-present

Text Mining of Naver News

• Search term: 메타버스 (Metaverse)
• Web scrapping using R (Naver API is available)
• Sentiment analysis and word network
• 10,677 after removing duplicates (in Jan. to Apr. 2023)
• 9,525 news after removing news less than 500 characters (n = 1,152
• min character = 500, max character = 13,634 per news

Pre-Process and Tokenize Texts (News Articles)

• R version 4.3.0

• Among at least 30 occurrences within each news

• At least .12 of correlations with “Cybersecurity”

• library(tidytext): unnest_tokens() to transform text data into a “tidy” format

• library(KoNLP): SimplePos09() to tokenize each into morphemes (the smallest meaningful unit)

• library(widyr): pairwise_cor() to find correlations of words associated with “Cybersecurity”

Word Network Analysis (“Cybersecurity”)

Code

suppressPackageStartupMessages({
  library(dplyr)
  library(tidygraph)
  library(ggraph)
  })

filter <- read.csv("https://raw.githubusercontent.com/mshin77/datasets/main/cybersecurity-workshop/filter.csv")

graph_meta <- filter %>%
    as_tbl_graph(directed = TRUE) %>%
    mutate(centrality = centrality_degree(mode = "out") / (44-1)) 

set.seed(2023)
layout_meta <- create_layout(graph_meta, layout = 'fr')

network_meta <- layout_meta %>% 
  ggraph() + 
  geom_edge_link(aes(edge_alpha = correlation, edge_width = correlation), alpha = 0.5, edge_colour = "skyblue") + 
  geom_node_point(aes(colour = centrality), size = 3) +
  geom_node_text(aes(label = name), size = 4, repel = TRUE) +
  scale_color_continuous(guide = 'legend', high = "#fa956b", low = "#ebc4b5") +
  theme_void(base_size = 6) +
  theme(legend.title=element_text(size = 11),
        legend.text=element_text(size = 11),
        legend.position = "right")

network_meta_g <- ggplot_build(network_meta)
network_meta_g$data[[3]] <- network_meta_g$data[[3]] %>%
  mutate(colour = if_else(label %in% c("Cybersecurity"), "red", "black"))
network_meta_gg <- ggplot_gtable(network_meta_g)

plot(network_meta_gg) 

Positive and Negative Words (“Disability” + “Cybersecurity”)

Code

suppressPackageStartupMessages({
  library(dplyr)
  library(ggplot2)
  library(tidytext)
})

sentiment_20 <- read.csv("https://raw.githubusercontent.com/mshin77/datasets/main/cybersecurity-workshop/sentiment_20.csv")

senti_plot <- sentiment_20 %>%
  mutate(word = reorder_within(word, n, sentiment)) %>%
  ggplot(aes(x = word, y = n, fill = sentiment)) +
  geom_col() +
  coord_flip() +
  geom_text(aes(label = n), hjust = -0.3) +
  facet_wrap(~ sentiment, scales = "free") +
  scale_x_reordered() +
  scale_y_continuous(expand = expansion(mult = c(0.05, 0.15))) +  
  labs(x = NULL) +
  theme_bw(base_size = 14) +
  theme(legend.title=element_text(size = 11),
        legend.text=element_text(size = 11),
        legend.position = "right")

senti_plot

Conclusion

• Advanced technology and its environments (e.g., metaverse) can support learners’ motivation and learning opportunities in various settings.

• Increased online and virtual opportunities and communications can also create new and unexpected security issues.

• We need to be proactive in utilizing and creating data and information.

• We need to consider limited access and resources for underrepresented groups.