Update README.md

README.md

@@ -1,5 +1,11 @@
 ---
 license: mit
+task_categories:
+- text2text-generation
+language:
+- en
+tags:
+- finance
 ---
 # Financial Parameter Database Overview
 
@@ -10,7 +16,13 @@ The Financial Parameter Database includes financial statements and parameters fr
 All text data, specifically the annual reports, originates from [NSE (National Stock Exchange of India)](https://www.nseindia.com/), where it is downloaded.Report have undergone pre-processing to eliminate stop words and special symbols, enhancing both clarity and efficiency in model training. The labels for the data, which consist of detailed financial data, are derived from [Moneycontrol](https://www.moneycontrol.com/).
 
 ## Data Description
-Each dataset encompasses a company's annual report. The labels for each dataset consist of specific financial parameter derived from this documents. This financial parameter includes, but is not limited to, Equity and Liabilities, Assets, Income, Expenses, and various financial ratios such as Earnings Per Share (EPS), Price-to-Earnings (P/E) ratio, and liquidity ratios.
+Each dataset encompasses a company's annual report. The labels for each dataset consist of specific financial parameters derived from these documents. These financial parameters include, but are not limited to, Equity and Liabilities, Assets, Income, Expenses, and various financial ratios such as Earnings Per Share (EPS), Price-to-Earnings (P/E) ratio, and liquidity ratios.
+
+Each dataset comprises three attributes: `CompanyName_Year`, `text`, and `label`:
+
+- `CompanyName_Year`: This attribute represents the company's symbol followed by the fiscal year, appearing in formats such as `3MINDIA_2021`. Here, `CompanyName_Year` combines the company's symbol and the corresponding year of the report.
+- `text`: This attribute contains the text of the company's annual report.
+- `label`: This attribute holds the financial parameters extracted from the annual report.
 
 ## Tokenization
 Initial tests indicate that approximately 98,304 tokens (16384*6) are necessary to tokenize a complete annual report, rendering this database both extensive and comprehensive for deep learning applications in financial text analysis.
@@ -110,4 +122,4 @@ The dataset is structured to facilitate training on advanced NLP models and is r
 - Enterprise Value (EV)
 - Market Cap to Revenue
 - Price to Book Value
-- Earnings Yield
+- Earnings Yield