update readme
Browse files
README.md
CHANGED
@@ -7,6 +7,7 @@ language:
|
|
7 |
- en
|
8 |
tags:
|
9 |
- keyphrase-generation
|
|
|
10 |
- paleontology
|
11 |
- astrophysics
|
12 |
- natural-language-processing
|
@@ -37,7 +38,21 @@ configs:
|
|
37 |
path: "astro/test.jsonl"
|
38 |
---
|
39 |
|
40 |
-
|
41 |
-
|
|
|
|
|
|
|
42 |
This dataset also includes three human-labeled test sets to assess the performance of keyphrase generation across these domains.
|
43 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
- en
|
8 |
tags:
|
9 |
- keyphrase-generation
|
10 |
+
- domain-adaptation
|
11 |
- paleontology
|
12 |
- astrophysics
|
13 |
- natural-language-processing
|
|
|
38 |
path: "astro/test.jsonl"
|
39 |
---
|
40 |
|
41 |
+
# `silk` synthetic training samples and human-labeled test sets for domain adaptation in keyphrase generation
|
42 |
+
|
43 |
+
This dataset contains the synthetic samples generated by 🧵 `silk`, a method that leverages citation contexts to create synthetic samples of documents paired with silver-standard keyphrases for adapting keyphrase generation models to new domains.
|
44 |
+
We applied `silk` on three domains: Natural Language Processing (nlp), Astrophysics (astro) and Paleontology (paleo).
|
45 |
+
|
46 |
This dataset also includes three human-labeled test sets to assess the performance of keyphrase generation across these domains.
|
47 |
|
48 |
+
## Citation
|
49 |
+
|
50 |
+
If you use this dataset, please cite the following paper:
|
51 |
+
|
52 |
+
```
|
53 |
+
Florian Boudin and Akiko Aizawa.
|
54 |
+
Unsupervised Domain Adaptation for Keyphrase Generation using Citation Context,
|
55 |
+
Proceedings of EMNLP 2024 (Findings).
|
56 |
+
```
|
57 |
+
|
58 |
+
|