Citation Network Visualization

Overview

The create_citation_network() function creates interactive visualizations showing how citations co-occur within documents. Citations appearing close together are connected, revealing patterns in how references are used and cited together.

Main Function

create_citation_network()

Create an interactive citation co-occurrence network.

Usage

create_citation_network(
  citation_analysis_results,
  max_distance = 800,
  min_connections = 2,
  show_labels = TRUE
)

Arguments

citation_analysis_results: Output from analyze_scientific_content()
max_distance: Maximum distance in characters between citations to create a connection
min_connections: Minimum number of connections required to include a node
show_labels: Logical. Whether to display citation labels on nodes

Value

An interactive visNetwork object with attributes:

stats: Network statistics including node count, edge count, distances, and section distribution

Basic Usage

Creating a Network

library(contentanalysis)
library(dplyr)

# First, analyze the document
doc <- pdf2txt_auto("paper.pdf", n_columns = 2)
analysis <- analyze_scientific_content(
  text = doc,
  doi = "10.xxxx/xxxxx",
  mailto = "your@email.com"
)

# Create network with default settings
network <- create_citation_network(
  citation_analysis_results = analysis,
  max_distance = 800,
  min_connections = 2,
  show_labels = TRUE
)

# Display the network
network

Network Statistics

Access detailed network statistics:

# Get statistics
stats <- attr(network, "stats")

# Basic network info
cat("Number of nodes:", stats$n_nodes, "\n")
cat("Number of edges:", stats$n_edges, "\n")
cat("Average distance:", round(stats$avg_distance), "characters\n")
cat("Maximum distance:", stats$max_distance, "characters\n")

# Section distribution
print(stats$section_distribution)

# Multi-section citations
if (nrow(stats$multi_section_citations) > 0) {
  cat("\nCitations appearing in multiple sections:\n")
  print(stats$multi_section_citations)
}

Visual Elements

Node Features

Size - Larger nodes have more connections - Size represents centrality in the citation network

Color - Indicates primary section where citation appears - Default color scheme: - Introduction: Light blue - Methods: Light green - Results: Light coral - Discussion: Light yellow - Abstract: Lavender

Border - Thicker borders (3px): Citations in multiple sections - Standard borders (1px): Citations in single section

Edge Features

Thickness - Thicker edges: Citations appearing closer together - Edge width inversely proportional to distance

Color - Red: Very close citations (≤300 characters) - Blue: Moderate distance (≤600 characters) - Gray: Distant citations (>600 characters)

Customization

Adjusting Distance Threshold

Control which citations are connected:

# Very close citations only
network_close <- create_citation_network(
  analysis,
  max_distance = 300,  # Within 300 characters
  min_connections = 1,
  show_labels = TRUE
)

# Moderate proximity
network_medium <- create_citation_network(
  analysis,
  max_distance = 600,
  min_connections = 2,
  show_labels = TRUE
)

# Broad connections
network_broad <- create_citation_network(
  analysis,
  max_distance = 1200,
  min_connections = 2,
  show_labels = TRUE
)

# Compare network sizes
cat("Close:", attr(network_close, "stats")$n_edges, "edges\n")
cat("Medium:", attr(network_medium, "stats")$n_edges, "edges\n")
cat("Broad:", attr(network_broad, "stats")$n_edges, "edges\n")

Filtering by Connections

Focus on well-connected citations:

# Include all connected citations
network_all <- create_citation_network(
  analysis,
  max_distance = 800,
  min_connections = 1
)

# Only "hub" citations (highly connected)
network_hubs <- create_citation_network(
  analysis,
  max_distance = 800,
  min_connections = 5,  # Must have 5+ connections
  show_labels = TRUE
)

# Compare
cat("All nodes:", attr(network_all, "stats")$n_nodes, "\n")
cat("Hub nodes:", attr(network_hubs, "stats")$n_nodes, "\n")

Label Display

Control label visibility:

# With labels (default for detailed inspection)
network_labeled <- create_citation_network(
  analysis,
  show_labels = TRUE
)

# Without labels (cleaner for presentations)
network_clean <- create_citation_network(
  analysis,
  show_labels = FALSE
)

Interpreting Networks

Identifying Patterns

# Get network statistics
stats <- attr(network, "stats")

# 1. Network density (how interconnected)
n_possible_edges <- stats$n_nodes * (stats$n_nodes - 1) / 2
density <- stats$n_edges / n_possible_edges
cat("Network density:", round(density, 3), "\n")

# 2. Find hub citations (top 25% by connections)
hub_threshold <- quantile(stats$section_distribution$n, 0.75)
hubs <- stats$section_distribution %>%
  filter(n >= hub_threshold) %>%
  arrange(desc(n))

cat("\nHub citations:\n")
print(hubs)

# 3. Section-specific patterns
section_summary <- stats$section_distribution %>%
  group_by(section) %>%
  summarise(
    n_citations = n(),
    avg_connections = mean(n),
    .groups = "drop"
  )

print(section_summary)

Citation Clusters

Identify groups of related citations:

# Citations with very close connections
close_pairs <- analysis$network_data %>%
  filter(distance < 200) %>%
  select(citation_from, citation_to, distance)

cat("Very close citation pairs:", nrow(close_pairs), "\n")
head(close_pairs)

# Find citation communities (simple approach)
# Citations that frequently co-occur
citation_freq <- analysis$network_data %>%
  count(citation_from, sort = TRUE)

top_cooccurring <- head(citation_freq, 10)
cat("\nMost frequently co-occurring citations:\n")
print(top_cooccurring)

Cross-Section Analysis

Citations appearing in multiple sections:

stats <- attr(network, "stats")

if (nrow(stats$multi_section_citations) > 0) {
  # Citations used across sections
  multi_section <- stats$multi_section_citations %>%
    arrange(desc(n_sections))
  
  cat("Citations in multiple sections:\n")
  print(multi_section)
  
  # These are often seminal or foundational works
  cat("\nThese citations appear in", 
      unique(multi_section$n_sections), 
      "different sections\n")
}

Advanced Analysis

Compare Networks Across Papers

# Analyze multiple papers
papers <- c("paper1.pdf", "paper2.pdf", "paper3.pdf")

networks <- list()
for (i in seq_along(papers)) {
  doc <- pdf2txt_auto(papers[i], n_columns = 2)
  analysis <- analyze_scientific_content(
    doc,
    doi = paste0("10.xxxx/", i),
    mailto = "your@email.com"
  )
  networks[[i]] <- create_citation_network(analysis)
}

# Compare network characteristics
comparison <- data.frame(
  paper = papers,
  nodes = sapply(networks, function(n) attr(n, "stats")$n_nodes),
  edges = sapply(networks, function(n) attr(n, "stats")$n_edges),
  avg_distance = sapply(networks, function(n) 
    round(attr(n, "stats")$avg_distance))
)

print(comparison)

# Calculate network densities
comparison$density <- comparison$edges / 
  (comparison$nodes * (comparison$nodes - 1) / 2)

print(comparison)

Network Metrics

Calculate advanced network metrics:

library(igraph)

# Convert to igraph object (if needed for advanced analysis)
# Note: You'll need the raw network data
edges <- analysis$network_data %>%
  filter(distance <= 800) %>%
  select(from = citation_from, to = citation_to, weight = distance)

g <- graph_from_data_frame(edges, directed = FALSE)

# Calculate metrics
metrics <- data.frame(
  citation = V(g)$name,
  degree = degree(g),
  betweenness = betweenness(g),
  closeness = closeness(g)
) %>%
  arrange(desc(degree))

cat("Top 10 citations by degree centrality:\n")
print(head(metrics, 10))

# Identify communities
communities <- cluster_louvain(g)
cat("\nNumber of communities:", length(communities), "\n")

Temporal Network Analysis

If years are available:

# Add year information to network
citation_years <- analysis$citation_references_mapping %>%
  select(citation_text_clean, cite_year) %>%
  distinct()

network_with_years <- analysis$network_data %>%
  left_join(citation_years, by = c("citation_from" = "citation_text_clean")) %>%
  rename(year_from = cite_year) %>%
  left_join(citation_years, by = c("citation_to" = "citation_text_clean")) %>%
  rename(year_to = cite_year)

# Analyze co-citation patterns by year
year_patterns <- network_with_years %>%
  filter(!is.na(year_from), !is.na(year_to)) %>%
  mutate(year_diff = abs(year_from - year_to)) %>%
  group_by(year_diff) %>%
  summarise(
    n_pairs = n(),
    avg_distance = mean(distance)
  )

cat("Co-citation patterns by year difference:\n")
print(year_patterns)

Export Network Data

Save Network Information

# Create export directory
dir.create("network_analysis", showWarnings = FALSE)

# 1. Network statistics
stats <- attr(network, "stats")
write.csv(stats$section_distribution,
          "network_analysis/section_distribution.csv",
          row.names = FALSE)

if (nrow(stats$multi_section_citations) > 0) {
  write.csv(stats$multi_section_citations,
            "network_analysis/multi_section_citations.csv",
            row.names = FALSE)
}

# 2. Edge list
write.csv(analysis$network_data,
          "network_analysis/edge_list.csv",
          row.names = FALSE)

# 3. Network summary
summary_data <- data.frame(
  metric = c("nodes", "edges", "density", "avg_distance", "max_distance"),
  value = c(
    stats$n_nodes,
    stats$n_edges,
    stats$n_edges / (stats$n_nodes * (stats$n_nodes - 1) / 2),
    stats$avg_distance,
    stats$max_distance
  )
)

write.csv(summary_data,
          "network_analysis/network_summary.csv",
          row.names = FALSE)

Save Interactive Network

# Save as HTML
library(htmlwidgets)

saveWidget(network, 
           "network_analysis/citation_network.html",
           selfcontained = TRUE)

cat("Network saved to: network_analysis/citation_network.html\n")

Use Cases

Use Case 1: Literature Review

Identify citation patterns in review papers:

# High connectivity expected
network <- create_citation_network(
  analysis,
  max_distance = 1000,  # Broader connections
  min_connections = 3
)

stats <- attr(network, "stats")

# Check network properties
cat("Network density:", 
    round(stats$n_edges / (stats$n_nodes * (stats$n_nodes - 1) / 2), 3),
    "\n")

# Hub citations (synthesis points)
hubs <- stats$section_distribution %>%
  filter(n >= quantile(n, 0.75))

cat("\nHub citations (potential synthesis points):\n")
print(hubs)

Use Case 2: Methods Paper

Focus on methodological citations:

# Filter to Methods section
methods_citations <- analysis$citations %>%
  filter(section == "Methods") %>%
  pull(citation_text_clean)

# Create network subset
methods_network_data <- analysis$network_data %>%
  filter(citation_from %in% methods_citations | 
         citation_to %in% methods_citations)

cat("Methods citations:", length(methods_citations), "\n")
cat("Methods citation pairs:", nrow(methods_network_data), "\n")

Use Case 3: Comparative Analysis

Compare citation networks across studies:

# Process multiple papers
papers_info <- data.frame(
  file = c("review.pdf", "empirical.pdf", "methods.pdf"),
  type = c("review", "empirical", "methods")
)

network_comparison <- data.frame()

for (i in 1:nrow(papers_info)) {
  doc <- pdf2txt_auto(papers_info$file[i], n_columns = 2)
  analysis <- analyze_scientific_content(doc, mailto = "your@email.com")
  net <- create_citation_network(analysis)
  stats <- attr(net, "stats")
  
  network_comparison <- rbind(network_comparison, data.frame(
    paper = papers_info$file[i],
    type = papers_info$type[i],
    nodes = stats$n_nodes,
    edges = stats$n_edges,
    density = stats$n_edges / (stats$n_nodes * (stats$n_nodes - 1) / 2),
    avg_distance = stats$avg_distance
  ))
}

print(network_comparison)

# Visualize comparison
library(ggplot2)
ggplot(network_comparison, aes(x = type, y = density, fill = type)) +
  geom_col() +
  labs(title = "Network Density by Paper Type",
       x = "Paper Type", y = "Network Density") +
  theme_minimal()

Tips and Best Practices

Choosing Parameters

max_distance: - 200-400: Very close citations (same paragraph) - 500-800: Moderate proximity (recommended) - 1000+: Broader connections (same section)

min_connections: - 1: Include all connected citations - 2-3: Filter isolated pairs (recommended) - 5+: Focus on hub citations only

Interpretation

Look for: - Clusters: Groups of related citations - Hubs: Frequently co-cited works - Bridges: Citations connecting different clusters - Isolates: Citations rarely co-cited

Performance

For large documents: - Start with higher min_connections - Use moderate max_distance - Consider section-specific networks - Save network as HTML for sharing

--- title: "Citation Network Visualization" --- ## Overview The `create_citation_network()` function creates interactive visualizations showing how citations co-occur within documents. Citations appearing close together are connected, revealing patterns in how references are used and cited together. ## Main Function ### create_citation_network() Create an interactive citation co-occurrence network. **Usage** ```r create_citation_network( citation_analysis_results, max_distance = 800, min_connections = 2, show_labels = TRUE ) ``` **Arguments** - `citation_analysis_results`: Output from `analyze_scientific_content()` - `max_distance`: Maximum distance in characters between citations to create a connection - `min_connections`: Minimum number of connections required to include a node - `show_labels`: Logical. Whether to display citation labels on nodes **Value** An interactive `visNetwork` object with attributes: - `stats`: Network statistics including node count, edge count, distances, and section distribution ## Basic Usage ### Creating a Network ```{r basic, eval=FALSE} library(contentanalysis) library(dplyr) # First, analyze the document doc <- pdf2txt_auto("paper.pdf", n_columns = 2) analysis <- analyze_scientific_content( text = doc, doi = "10.xxxx/xxxxx", mailto = "your@email.com" ) # Create network with default settings network <- create_citation_network( citation_analysis_results = analysis, max_distance = 800, min_connections = 2, show_labels = TRUE ) # Display the network network ``` ### Network Statistics Access detailed network statistics: ```{r stats, eval=FALSE} # Get statistics stats <- attr(network, "stats") # Basic network info cat("Number of nodes:", stats$n_nodes, "\n") cat("Number of edges:", stats$n_edges, "\n") cat("Average distance:", round(stats$avg_distance), "characters\n") cat("Maximum distance:", stats$max_distance, "characters\n") # Section distribution print(stats$section_distribution) # Multi-section citations if (nrow(stats$multi_section_citations) > 0) { cat("\nCitations appearing in multiple sections:\n") print(stats$multi_section_citations) } ``` ## Visual Elements ### Node Features **Size** - Larger nodes have more connections - Size represents centrality in the citation network **Color** - Indicates primary section where citation appears - Default color scheme: - Introduction: Light blue - Methods: Light green - Results: Light coral - Discussion: Light yellow - Abstract: Lavender **Border** - Thicker borders (3px): Citations in multiple sections - Standard borders (1px): Citations in single section ### Edge Features **Thickness** - Thicker edges: Citations appearing closer together - Edge width inversely proportional to distance **Color** - Red: Very close citations (≤300 characters) - Blue: Moderate distance (≤600 characters) - Gray: Distant citations (>600 characters) ## Customization ### Adjusting Distance Threshold Control which citations are connected: ```{r distance, eval=FALSE} # Very close citations only network_close <- create_citation_network( analysis, max_distance = 300, # Within 300 characters min_connections = 1, show_labels = TRUE ) # Moderate proximity network_medium <- create_citation_network( analysis, max_distance = 600, min_connections = 2, show_labels = TRUE ) # Broad connections network_broad <- create_citation_network( analysis, max_distance = 1200, min_connections = 2, show_labels = TRUE ) # Compare network sizes cat("Close:", attr(network_close, "stats")$n_edges, "edges\n") cat("Medium:", attr(network_medium, "stats")$n_edges, "edges\n") cat("Broad:", attr(network_broad, "stats")$n_edges, "edges\n") ``` ### Filtering by Connections Focus on well-connected citations: ```{r connections, eval=FALSE} # Include all connected citations network_all <- create_citation_network( analysis, max_distance = 800, min_connections = 1 ) # Only "hub" citations (highly connected) network_hubs <- create_citation_network( analysis, max_distance = 800, min_connections = 5, # Must have 5+ connections show_labels = TRUE ) # Compare cat("All nodes:", attr(network_all, "stats")$n_nodes, "\n") cat("Hub nodes:", attr(network_hubs, "stats")$n_nodes, "\n") ``` ### Label Display Control label visibility: ```{r labels, eval=FALSE} # With labels (default for detailed inspection) network_labeled <- create_citation_network( analysis, show_labels = TRUE ) # Without labels (cleaner for presentations) network_clean <- create_citation_network( analysis, show_labels = FALSE ) ``` ## Interpreting Networks ### Identifying Patterns ```{r patterns, eval=FALSE} # Get network statistics stats <- attr(network, "stats") # 1. Network density (how interconnected) n_possible_edges <- stats$n_nodes * (stats$n_nodes - 1) / 2 density <- stats$n_edges / n_possible_edges cat("Network density:", round(density, 3), "\n") # 2. Find hub citations (top 25% by connections) hub_threshold <- quantile(stats$section_distribution$n, 0.75) hubs <- stats$section_distribution %>% filter(n >= hub_threshold) %>% arrange(desc(n)) cat("\nHub citations:\n") print(hubs) # 3. Section-specific patterns section_summary <- stats$section_distribution %>% group_by(section) %>% summarise( n_citations = n(), avg_connections = mean(n), .groups = "drop" ) print(section_summary) ``` ### Citation Clusters Identify groups of related citations: ```{r clusters, eval=FALSE} # Citations with very close connections close_pairs <- analysis$network_data %>% filter(distance < 200) %>% select(citation_from, citation_to, distance) cat("Very close citation pairs:", nrow(close_pairs), "\n") head(close_pairs) # Find citation communities (simple approach) # Citations that frequently co-occur citation_freq <- analysis$network_data %>% count(citation_from, sort = TRUE) top_cooccurring <- head(citation_freq, 10) cat("\nMost frequently co-occurring citations:\n") print(top_cooccurring) ``` ### Cross-Section Analysis Citations appearing in multiple sections: ```{r cross_section, eval=FALSE} stats <- attr(network, "stats") if (nrow(stats$multi_section_citations) > 0) { # Citations used across sections multi_section <- stats$multi_section_citations %>% arrange(desc(n_sections)) cat("Citations in multiple sections:\n") print(multi_section) # These are often seminal or foundational works cat("\nThese citations appear in", unique(multi_section$n_sections), "different sections\n") } ``` ## Advanced Analysis ### Compare Networks Across Papers ```{r compare, eval=FALSE} # Analyze multiple papers papers <- c("paper1.pdf", "paper2.pdf", "paper3.pdf") networks <- list() for (i in seq_along(papers)) { doc <- pdf2txt_auto(papers[i], n_columns = 2) analysis <- analyze_scientific_content( doc, doi = paste0("10.xxxx/", i), mailto = "your@email.com" ) networks[[i]] <- create_citation_network(analysis) } # Compare network characteristics comparison <- data.frame( paper = papers, nodes = sapply(networks, function(n) attr(n, "stats")$n_nodes), edges = sapply(networks, function(n) attr(n, "stats")$n_edges), avg_distance = sapply(networks, function(n) round(attr(n, "stats")$avg_distance)) ) print(comparison) # Calculate network densities comparison$density <- comparison$edges / (comparison$nodes * (comparison$nodes - 1) / 2) print(comparison) ``` ### Network Metrics Calculate advanced network metrics: ```{r metrics, eval=FALSE} library(igraph) # Convert to igraph object (if needed for advanced analysis) # Note: You'll need the raw network data edges <- analysis$network_data %>% filter(distance <= 800) %>% select(from = citation_from, to = citation_to, weight = distance) g <- graph_from_data_frame(edges, directed = FALSE) # Calculate metrics metrics <- data.frame( citation = V(g)$name, degree = degree(g), betweenness = betweenness(g), closeness = closeness(g) ) %>% arrange(desc(degree)) cat("Top 10 citations by degree centrality:\n") print(head(metrics, 10)) # Identify communities communities <- cluster_louvain(g) cat("\nNumber of communities:", length(communities), "\n") ``` ### Temporal Network Analysis If years are available: ```{r temporal, eval=FALSE} # Add year information to network citation_years <- analysis$citation_references_mapping %>% select(citation_text_clean, cite_year) %>% distinct() network_with_years <- analysis$network_data %>% left_join(citation_years, by = c("citation_from" = "citation_text_clean")) %>% rename(year_from = cite_year) %>% left_join(citation_years, by = c("citation_to" = "citation_text_clean")) %>% rename(year_to = cite_year) # Analyze co-citation patterns by year year_patterns <- network_with_years %>% filter(!is.na(year_from), !is.na(year_to)) %>% mutate(year_diff = abs(year_from - year_to)) %>% group_by(year_diff) %>% summarise( n_pairs = n(), avg_distance = mean(distance) ) cat("Co-citation patterns by year difference:\n") print(year_patterns) ``` ## Export Network Data ### Save Network Information ```{r export, eval=FALSE} # Create export directory dir.create("network_analysis", showWarnings = FALSE) # 1. Network statistics stats <- attr(network, "stats") write.csv(stats$section_distribution, "network_analysis/section_distribution.csv", row.names = FALSE) if (nrow(stats$multi_section_citations) > 0) { write.csv(stats$multi_section_citations, "network_analysis/multi_section_citations.csv", row.names = FALSE) } # 2. Edge list write.csv(analysis$network_data, "network_analysis/edge_list.csv", row.names = FALSE) # 3. Network summary summary_data <- data.frame( metric = c("nodes", "edges", "density", "avg_distance", "max_distance"), value = c( stats$n_nodes, stats$n_edges, stats$n_edges / (stats$n_nodes * (stats$n_nodes - 1) / 2), stats$avg_distance, stats$max_distance ) ) write.csv(summary_data, "network_analysis/network_summary.csv", row.names = FALSE) ``` ### Save Interactive Network ```{r save_network, eval=FALSE} # Save as HTML library(htmlwidgets) saveWidget(network, "network_analysis/citation_network.html", selfcontained = TRUE) cat("Network saved to: network_analysis/citation_network.html\n") ``` ## Use Cases ### Use Case 1: Literature Review Identify citation patterns in review papers: ```{r usecase1, eval=FALSE} # High connectivity expected network <- create_citation_network( analysis, max_distance = 1000, # Broader connections min_connections = 3 ) stats <- attr(network, "stats") # Check network properties cat("Network density:", round(stats$n_edges / (stats$n_nodes * (stats$n_nodes - 1) / 2), 3), "\n") # Hub citations (synthesis points) hubs <- stats$section_distribution %>% filter(n >= quantile(n, 0.75)) cat("\nHub citations (potential synthesis points):\n") print(hubs) ``` ### Use Case 2: Methods Paper Focus on methodological citations: ```{r usecase2, eval=FALSE} # Filter to Methods section methods_citations <- analysis$citations %>% filter(section == "Methods") %>% pull(citation_text_clean) # Create network subset methods_network_data <- analysis$network_data %>% filter(citation_from %in% methods_citations | citation_to %in% methods_citations) cat("Methods citations:", length(methods_citations), "\n") cat("Methods citation pairs:", nrow(methods_network_data), "\n") ``` ### Use Case 3: Comparative Analysis Compare citation networks across studies: ```{r usecase3, eval=FALSE} # Process multiple papers papers_info <- data.frame( file = c("review.pdf", "empirical.pdf", "methods.pdf"), type = c("review", "empirical", "methods") ) network_comparison <- data.frame() for (i in 1:nrow(papers_info)) { doc <- pdf2txt_auto(papers_info$file[i], n_columns = 2) analysis <- analyze_scientific_content(doc, mailto = "your@email.com") net <- create_citation_network(analysis) stats <- attr(net, "stats") network_comparison <- rbind(network_comparison, data.frame( paper = papers_info$file[i], type = papers_info$type[i], nodes = stats$n_nodes, edges = stats$n_edges, density = stats$n_edges / (stats$n_nodes * (stats$n_nodes - 1) / 2), avg_distance = stats$avg_distance )) } print(network_comparison) # Visualize comparison library(ggplot2) ggplot(network_comparison, aes(x = type, y = density, fill = type)) + geom_col() + labs(title = "Network Density by Paper Type", x = "Paper Type", y = "Network Density") + theme_minimal() ``` ## Tips and Best Practices ::: {.callout-tip} ## Choosing Parameters **max_distance**: - 200-400: Very close citations (same paragraph) - 500-800: Moderate proximity (recommended) - 1000+: Broader connections (same section) **min_connections**: - 1: Include all connected citations - 2-3: Filter isolated pairs (recommended) - 5+: Focus on hub citations only ::: ::: {.callout-tip} ## Interpretation Look for: - **Clusters**: Groups of related citations - **Hubs**: Frequently co-cited works - **Bridges**: Citations connecting different clusters - **Isolates**: Citations rarely co-cited ::: ::: {.callout-note} ## Performance For large documents: - Start with higher `min_connections` - Use moderate `max_distance` - Consider section-specific networks - Save network as HTML for sharing ::: ## See Also - [Citation Analysis](citation-analysis.qmd): Extract and analyze citations - [Content Analysis](content-analysis.qmd): Main analysis function - [Tutorial](../tutorial.qmd): Complete workflow examples