sumitranjan
commited on
Commit
•
849d4a7
1
Parent(s):
581133b
Update README.md
Browse files
README.md
CHANGED
@@ -6,6 +6,15 @@ widget:
|
|
6 |
- text: "Company went through great loss due to lawsuit in Q1"
|
7 |
---
|
8 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
9 |
## What is Earning Call Transcript?
|
10 |
|
11 |
An earnings call is a teleconference, or webcast, in which a public company discusses the financial results of a reporting period. The name comes from earnings per share, the bottom line number in the income statement divided by the number of shares outstanding.
|
@@ -16,14 +25,6 @@ Scraped 10 years of earning call transcript data for 10 companies like Apple, go
|
|
16 |
|
17 |
And then used Loughran-McDonald sentiment lexicon and Use FinancialPhraseBank [Malo, P., Sinha, A., Korhonen, P., Wallenius, J., & Takala, P. (2014). Good debt or bad debt: Detecting semantic orientations in economic texts. Journal of the Association for Information Science and Technology, 65(4), 782-796.] for data annotation.
|
18 |
|
19 |
-
## What is RoBERTa
|
20 |
-
|
21 |
-
RoBERTa builds on BERT’s language masking strategy and modifies key hyperparameters in BERT, including removing BERT’s next-sentence pretraining objective, and training with much larger mini-batches and learning rates. RoBERTa was also trained on an order of magnitude more data than BERT, for a longer amount of time. This allows RoBERTa representations to generalize even better to downstream tasks compared to BERT.
|
22 |
-
|
23 |
-
## What is Roberta-Earning-Call-Transcript-Classification?
|
24 |
-
|
25 |
-
Roberta-Earning-Call-Transcript-Classification is a Multi-Label Classification Model trained with Annotated earning call transcript data. This model could be very helpful in finding Negative, Positive, Litigious, Constraining and Uncertain thing in the sentence. This could be really helpful in analyzing Profit warning of a company.
|
26 |
-
|
27 |
## Hyperparameters
|
28 |
|
29 |
| Parameter | |
|
|
|
6 |
- text: "Company went through great loss due to lawsuit in Q1"
|
7 |
---
|
8 |
|
9 |
+
|
10 |
+
## What is RoBERTa
|
11 |
+
|
12 |
+
RoBERTa builds on BERT’s language masking strategy and modifies key hyperparameters in BERT, including removing BERT’s next-sentence pretraining objective, and training with much larger mini-batches and learning rates. RoBERTa was also trained on an order of magnitude more data than BERT, for a longer amount of time. This allows RoBERTa representations to generalize even better to downstream tasks compared to BERT.
|
13 |
+
|
14 |
+
## What is Roberta-Earning-Call-Transcript-Classification?
|
15 |
+
|
16 |
+
Roberta-Earning-Call-Transcript-Classification is a Multi-Label Classification Model trained with Annotated earning call transcript data. This model could be very helpful in finding Negative, Positive, Litigious, Constraining and Uncertain thing in the sentence. This could be really helpful in analyzing Profit warning of a company.
|
17 |
+
|
18 |
## What is Earning Call Transcript?
|
19 |
|
20 |
An earnings call is a teleconference, or webcast, in which a public company discusses the financial results of a reporting period. The name comes from earnings per share, the bottom line number in the income statement divided by the number of shares outstanding.
|
|
|
25 |
|
26 |
And then used Loughran-McDonald sentiment lexicon and Use FinancialPhraseBank [Malo, P., Sinha, A., Korhonen, P., Wallenius, J., & Takala, P. (2014). Good debt or bad debt: Detecting semantic orientations in economic texts. Journal of the Association for Information Science and Technology, 65(4), 782-796.] for data annotation.
|
27 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
28 |
## Hyperparameters
|
29 |
|
30 |
| Parameter | |
|