-
Notifications
You must be signed in to change notification settings - Fork 1
/
info.json
23 lines (23 loc) · 1.14 KB
/
info.json
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
{
"abstract": "We introduce Refinery, an open source platform for exploring large text document collections with topic models. Refinery is a standalone web application driven by a graphical interface, so it is usable by those without machine learning or programming expertise. Users can interactively organize articles by topic and also refine this organization with phrase-level analysis. Under the hood, we train Bayesian nonparametric topic models that can adapt model complexity to the provided data with scalable learning algorithms. <a href=\"http://daeilkim.github.io/refinery/\">The project website</a> contains Python code and further documentation.",
"authors": [
"Daeil Kim",
"Benjamin F. Swanson",
"Michael C. Hughes",
"Erik B. Sudderth"
],
"id": "15-441",
"issue": 12,
"pages": [
1,
5
],
"title": "Refinery: An Open Source Topic Modeling Web Platform",
"volume": 18,
"year": 2017,
"extra_links": [
["code", "https://github.com/daeilkim/refinery"],
["webpage", "http://daeilkim.github.io/refinery/"]
],
"special_issue": "MLOSS"
}