Analyzing Feeds on Blue Sky

On Blue Sky users have the ability to create custom feeds based on specific keywords. These feeds aggregate content, for instance, a user might create a feed around the hashtag #rstats to gather all relevant content about. Let’s delve into the dynamics of such feeds created by users.

Load the package

library(atrrr)

Retrieving a Feed

Our starting point is to extract the posts from a feed. We’re focusing on a feed curated by “andrew.heiss.phd”.

# Fetching the feed posts
feeds <- get_feeds_created_by(actor = "andrew.heiss.phd") |>
  dplyr::glimpse()
#> Rows: 4
#> Columns: 20
#> $ uri                       <chr> "at://did:plc:2zcfjz…
#> $ cid                       <chr> "bafyreicvjczzxxhrkr…
#> $ did                       <chr> "did:web:skyfeed.me"…
#> $ creator_did               <chr> "did:plc:2zcfjzyocp6…
#> $ creator_handle            <chr> "andrew.heiss.phd", …
#> $ creator_displayName       <chr> "Andrew Heiss (🎄fes…
#> $ creator_description       <chr> "Assistant professor…
#> $ creator_avatar            <chr> "https://cdn.bsky.ap…
#> $ creator_indexedAt         <chr> "2023-11-26T17:07:49…
#> $ creator_viewer_muted      <lgl> FALSE, FALSE, FALSE,…
#> $ creator_viewer_blockedBy  <lgl> FALSE, FALSE, FALSE,…
#> $ creator_viewer_following  <chr> "at://did:plc:ntd53a…
#> $ creator_viewer_followedBy <chr> "at://did:plc:2zcfjz…
#> $ displayName               <chr> "Public Admin/Policy…
#> $ description               <chr> "A feed for public a…
#> $ avatar                    <chr> "https://cdn.bsky.ap…
#> $ likeCount                 <int> 80, 19, 0, 94
#> $ indexedAt                 <chr> "2023-09-21T01:37:55…
#> $ created_at                <dttm> 2023-09-21 01:37:55,…
#> $ viewer_like               <chr> NA, NA, NA, "at://d…

# Filtering for a specific keyword, for example "#rstats"
rstat_feed <- feeds |>
  filter(displayName == "#rstats")

# Extracting posts from this curated feed
rstat_posts <- get_feed(rstat_feed$uri, limit = 200) |>
  dplyr::glimpse()
#> Rows: 79
#> Columns: 15
#> $ uri           <chr> "at://did:plc:kizxn77jkp4p5vqzap…
#> $ cid           <chr> "bafyreiaabdszn3z7ka5c5zm2e2ufkj…
#> $ author_handle <chr> "pmaier1971.bsky.social", "pmaie…
#> $ author_name   <chr> "Philipp Maier", "Philipp Maier"…
#> $ text          <chr> "US #HouseholdDebt in Jul 2023 a…
#> $ author_data   <list> ["did:plc:kizxn77jkp4p5vqzapbed…
#> $ post_data     <list> ["US #HouseholdDebt in Jul 2023…
#> $ embed_data    <list> ["app.bsky.embed.images#view", …
#> $ reply_count   <int> 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0,…
#> $ repost_count  <int> 0, 0, 0, 2, 0, 0, 4, 3, 0, 0, 0,…
#> $ like_count    <int> 0, 0, 7, 2, 5, 11, 2, 9, 0, 0, 0…
#> $ indexed_at    <dttm> 2024-01-02 13:17:27, 2024-01-02…
#> $ in_reply_to   <chr> NA, NA, NA, NA, NA, NA, NA, NA, …
#> $ in_reply_root <chr> NA, NA, NA, NA, NA, NA, NA, NA, …
#> $ quotes        <chr> NA, NA, NA, NA, NA, NA, NA, NA, …

Identifying Top Contributors

Who are the leading voices within a particular topic? This analysis highlights users who are frequently contributing to the #rstats feed.

library(ggplot2)

# Identifying the top 10 contributors
rstat_posts |>
  count(handle = author_handle, sort = T) |>
  slice(1:10) |>
  mutate(handle = forcats::fct_reorder(handle, n)) |>
  ggplot(aes(handle, n)) +
  geom_col() +
  coord_flip() +
  theme_minimal()

Top 10 #rstats contributors

Recognizing Influential Voices

Volume doesn’t always translate to influence. Some users may post less frequently but their contributions resonate deeply with the community.

# Identifying top 10 influential voices based on likes
rstat_posts |>
  group_by(author_handle) |>
  summarize(like_count = sum(like_count)) |>
  ungroup() |>
  arrange(desc(like_count)) |>
  slice(1:10) |>
  mutate(handle = forcats::fct_reorder(author_handle, like_count)) |>
  ggplot(aes(handle, like_count)) +
  geom_col() +
  coord_flip() +
  theme_minimal()

Top 10 #rstats contributors based on likes

Most Famous #rstats skeet