quim-motger
commited on
Commit
•
27c8aec
1
Parent(s):
7037c51
Update README.md
Browse files
README.md
CHANGED
@@ -28,4 +28,29 @@ widget:
|
|
28 |
example_title: "Example 8"
|
29 |
- text: "There is no setting for recurring alarms"
|
30 |
example_title: "Example 9"
|
31 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
28 |
example_title: "Example 8"
|
29 |
- text: "There is no setting for recurring alarms"
|
30 |
example_title: "Example 9"
|
31 |
+
---
|
32 |
+
|
33 |
+
# T-FREX BERT base model (uncased)
|
34 |
+
|
35 |
+
T-FREX is a transformer-based feature extraction method for mobile app reviews based on fine-tuning Large Language Models (LLMs) for a named entity recognition task. We collect a dataset of ground truth features from users in a real crowdsourced software recommendation platform, and we use this dataset to fine-tune multiple LLMs under different data configurations. We assess the performance of T-FREX with respect to this ground truth, and we complement our analysis by comparing T-FREX with a baseline method from the field. Finally, we assess the quality of new features predicted by T-FREX through an external human evaluation. Results show that T-FREX outperforms on average the traditional syntactic-based method, especially when discovering new features from a domain for which the model has been fine-tuned.
|
36 |
+
|
37 |
+
Source code for data generation, fine-tuning and model inference are available in the original [GitHub repository](https://github.com/gessi-chatbots/t-frex/).
|
38 |
+
|
39 |
+
## Model description
|
40 |
+
|
41 |
+
This version of T-FREX has been fine-tuned for [token classification](https://huggingface.co/docs/transformers/tasks/token_classification#train) from [BERT base model (uncased)](https://huggingface.co/bert-base-uncased).
|
42 |
+
|
43 |
+
## Model variations
|
44 |
+
|
45 |
+
T-FREX includes a set of released, fine-tuned models which are compared in the original study (to be published).
|
46 |
+
|
47 |
+
- [**t-frex-bert-base-uncased**](https://huggingface.co/quim-motger/t-frex-bert-base-uncased)
|
48 |
+
- [**t-frex-bert-large-uncased**](https://huggingface.co/quim-motger/t-frex-bert-large-uncased)
|
49 |
+
- [**t-frex-roberta-base**](https://huggingface.co/quim-motger/t-frex-roberta-base)
|
50 |
+
- [**t-frex-roberta-large**](https://huggingface.co/quim-motger/t-frex-roberta-large)
|
51 |
+
- [**t-frex-xlnet-base-cased**](https://huggingface.co/quim-motger/t-frex-xlnet-base-cased)
|
52 |
+
- [**t-frex-xlnet-large-cased**](https://huggingface.co/quim-motger/t-frex-xlnet-large-cased)
|
53 |
+
|
54 |
+
## How to use
|
55 |
+
|
56 |
+
You can use this model following the instructions for [model inference for token classification](https://huggingface.co/docs/transformers/tasks/token_classification#inference).
|