Stack Exchange — ai.stackexchange.com

The content of ai.stackexchange.com (https://ai.stackexchange.com/), a Stack Exchange question-and-answer site, as published in the Stack Exchange Data Dump of April 2024 on the Internet Archive. The posts table holds 56,113 rows: 26,764 posts (12,380 questions, 12,460 answers, 954 tag wiki excerpts, 954 tag wikis, 13 posts of type 6 and 3 posts of type 7) and 29,349 comments, written between 2 August 2016 and 31 March 2024. Beside it are 71,811 user accounts, 93,175 votes, 62,748 badges awarded, 986 tags, 105,993 post revisions and 1,792 links between posts. Every answer keeps the question it answers and every comment the post it was written on, so the threads can be rebuilt. The Internet Archive item describes the dump as anonymised; the votes table names the voter in only 307 rows.

#Artificial intelligence#Expertise network#Reputation#Stack exchange data dump

Download data package README Original files as received BibTeX JSONSimilar datasets

Analytics

The interactive network needs JavaScript.

· graph.json

The full graph has 6,708 nodes; this view shows the 1,000-node sample the pipeline drew for display. Node size is centrality, colour is community.

6.7K
Nodes
10.4K
Edges
2.31e-04
Density
263
Components
-0.09228
Assortativity
0.4%
Reciprocity
4K
Bridges
Most central nodes
80.0161
187580.01282
24440.00701
16710.00656
41990.00656
226590.00552
165650.00522
168710.00447
253620.00447
28440.00447
Content metrics

56,113 records scored · min–maxmean ± 0.675 σ (the interquartile range if the values are roughly normal)mean

x scoreμ 1.21 ± 3.295
min -6max 182
x view countμ 867.192 ± 3.6K
min 4max 119.6K
x answer countμ 1.007 ± 1.032
min 0max 19
x comment countμ 1.097 ± 1.995
min 0max 34
x favorite count1.0% true
true 1.0%false 99.0%
10.2
Flesch–Kincaid grade
US school grade level
10.9
Automated readability index
Frequent terms
modellearningdatanetworkusefunctionliketrainingneuralstatequestionproblemtimeinputexamplevalueoutputpolicydifferentactionlossselfimageanswerneedlayerwantdonthinknumberknowwaycasealgorithmsetagentbasednetworksrewardvaluesgoodpossiblerightpaperlayersgiventrainimagesmeanfind
Topic clusters

Topic 0

question
answer
thank
post
ask
edit
paper
link

Topic 1

function
network
layer
neural
gradient
loss
output
weight

Topic 2

datum
model
set
training
test
dataset
validation
train

Topic 3

word
language
model
transformer
text
attention
token
sentence

Topic 4

human
intelligence
artificial
system
agi
computer
intelligent
machine

Topic 5

policy
state
action
reward
value
function
mdp
probability

Topic 6

game
play
player
win
chess
card
tree
board

Topic 7

object
image
detection
box
segmentation
detect
face
bounding

Topic 8

image
layer
convolution
cnn
convolutional
filter
channel
feature

Topic 9

agent
reinforcement
learning
action
reward
state
environment
policy

Topic 10

gan
discriminator
generator
image
vae
adversarial
latent
generate
Top hashtags
#1105
#271
#print68
#model50
#336
#include29
#page21
#get15
#end14
#data11
#e11
#if11
#hidden10
#train10
#x10
#49
#and9
#self9
#torch9
#assert8

Computed over the full corpus.

Record

Catalog key
stackexchange-ai-stackexchange-com
Authors
Stack Exchange, Inc.
Year
2024
Source paper
Stack Exchange Data Dump
License
CC BY-SA 3.0 / 4.0 (by contribution date) (CC-BY-SA-3.0 AND CC-BY-SA-4.0)
Records
67,817
Nodes
6,708
Edges
10,407 (directed)
Nodes are
user (Stack Exchange account)
Edges are
answer collaboration (A answered B's question)
Attributes
nodes have attributes; edges have attributes
Text corpus
question titles, post bodies, comment text, tag names
Languages
en
Collected
2016/2024 (2010s, 2020s)
Tasks
From the graph: link prediction, community detection
Region
Global (inferred: the platform is used worldwide)

Provenance

Originally published at
https://archive.org/details/stackexchange link checked
Original files as received
The source files, unconverted, kept for reference: mirrored on echo.directory on 2026-10-09, 35.3 MB across 1 files. Checksums in MANIFEST.json. For use, start with the data package below.
Data package
Version 1.0.0 of this dataset as an ECHO Data Package: posts (56,113 rows), users (71,811 rows), x_votes (93,175 rows), x_badges (62,748 rows), x_tags (986 rows), x_post_history (105,993 rows), x_post_links (1,792 rows).
Structure
Temporal (timestamped), Node attributes, Text, Threaded conversations

How to cite

BibTeX

@misc{echo2026,
  title  = {ECHO: Exploratory Compendium of Humanity Online},
  author = {Switzer, Braden and Zafarani, Reza},
  year   = {2026},
  url    = {https://echo.directory},
  note   = {Syracuse University}
}

@misc{stackexchange_2024,
  title     = {Stack Exchange Data Dump},
  author    = {{Stack Exchange, Inc.}},
  year      = {2024},
  month     = {apr},
  publisher = {Internet Archive},
  url       = {https://archive.org/details/stackexchange}
}

APA

Switzer, B., & Zafarani, R. (2026). ECHO: Exploratory Compendium of Humanity Online [Data repository]. Syracuse University. https://echo.directory

Stack Exchange, Inc. (2024). Stack Exchange — ai.stackexchange.com [Data set]. https://stackexchange-ai.echo.directory/

MLA

Switzer, Braden, and Reza Zafarani. ECHO: Exploratory Compendium of Humanity Online. Syracuse University, 2026, echo.directory.

Stack Exchange, Inc. Stack Exchange — ai.stackexchange.com. 2024, stackexchange-ai.echo.directory/.

BibTeX · RIS · CSL-JSON

Similar datasets 10 all with scores →