mch-dd commited on
Commit
b09d02e
1 Parent(s): 1fda0bb

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +110 -0
README.md ADDED
@@ -0,0 +1,110 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc0-1.0
3
+ task_categories:
4
+ - text-generation
5
+ pretty_name: SEC Annual Reports (Form 10-K) 1993 - 2024
6
+ ---
7
+
8
+ ## SEC Annual Reports (Form 10-K) 1993-2024
9
+
10
+ ### Dataset Overview
11
+
12
+ This dataset comprises SEC annual reports (Form 10-K) for the years 1993 to 2024, providing comprehensive coverage of publicly traded companies' financial and business information. The reports are stored in Parquet format, ensuring efficient storage and quick access. This dataset was meticulously compiled using the EDGAR-Crawler toolkit, which facilitates the extraction and processing of SEC filings from the EDGAR database.
13
+
14
+ ### Dataset Structure
15
+
16
+ #### Data Files
17
+
18
+ The dataset is organized into separate Parquet files for each year, making it easy to navigate and utilize:
19
+
20
+ - 1993.parquet
21
+ - 1994.parquet
22
+ - 1995.parquet
23
+ - 1996.parquet
24
+ - 1997.parquet
25
+ - 1998.parquet
26
+ - 1999.parquet
27
+ - 2000.parquet
28
+ - 2001.parquet
29
+ - 2002.parquet
30
+ - 2003.parquet
31
+ - 2004.parquet
32
+ - 2005.parquet
33
+ - 2006.parquet
34
+ - 2007.parquet
35
+ - 2008.parquet
36
+ - 2009.parquet
37
+ - 2010.parquet
38
+ - 2011.parquet
39
+ - 2012.parquet
40
+ - 2013.parquet
41
+ - 2014.parquet
42
+ - 2015.parquet
43
+ - 2016.parquet
44
+ - 2017.parquet
45
+ - 2018.parquet
46
+ - 2019.parquet
47
+ - 2020.parquet
48
+ - 2021.parquet
49
+ - 2022.parquet
50
+ - 2023.parquet
51
+ - 2024.parquet
52
+
53
+ ### Summary Statistics
54
+
55
+ Across these years, the dataset encapsulates a total of 7,245,966,226 words spread over 245,211 entries, with an average of 34,324.36 words per entry. Notably, there are 4,043 documents with zero words, reflecting the occasional nature of filings that contain no textual content.
56
+
57
+ ### Fields Included
58
+
59
+ Each Parquet file contains detailed fields that provide a comprehensive view of each report:
60
+
61
+ - `filename`: The filename of the filing document (e.g., "1089297_21929025_2004.htm").
62
+ - `id`: A unique identifier for the filing, formatted as "cik_year" (e.g., "1089297_2021").
63
+ - `year`: The year of the filing.
64
+ - `cik`: The Central Index Key assigned to the company (e.g., "1089297").
65
+ - `text`: The full text of the filing.
66
+ - `word_count`: The number of words in the filing text.
67
+ - `character_count`: The number of characters in the filing text.
68
+
69
+ ### Source and Methodology
70
+
71
+ #### Source
72
+
73
+ - Until 2020: The data have been collected from https://zenodo.org/records/5528490.
74
+ - From 2021 to 2024: The data were collected using the EDGAR-Crawler toolkit, which facilitated the extraction and processing of SEC filings from the EDGAR database.
75
+
76
+ #### Methodology
77
+
78
+ 1. **Crawling**: The EDGAR-Crawler toolkit was utilized to download the 10-K filings for each specified year.
79
+ 2. **Extraction and Cleaning**: The filings were extracted and cleaned to ensure a structured and clean dataset.
80
+ 3. **Integration**: This dataset is seamlessly integrated with existing datasets from 1993 to 2020, providing a continuous and comprehensive record of SEC annual reports for extensive research and analysis.
81
+
82
+ ### Use Cases
83
+
84
+ This dataset is invaluable for various applications, including but not limited to:
85
+
86
+ - **Academic Research**: Researchers in economics, finance, and business management can leverage this dataset to conduct detailed and expansive analyses, enhancing the scope and depth of their studies with robust financial data.
87
+ - **Financial Analysis**: Professionals in finance can utilize the detailed reports to bolster financial analysis, strategic planning, and decision-making processes, ensuring well-informed and data-driven insights.
88
+ - **NLP Applications**: The structured textual data in this dataset supports natural language processing (NLP) research and applications, enabling the development of advanced models and tools for financial document analysis and more.
89
+
90
+ ### General Dataset Statistics
91
+
92
+ - **Total number of words**: 7,245,966,226
93
+ - **Total number of entries**: 245,211
94
+ - **Average number of words per entry**: 34,324.36
95
+ - **Number of zero-word documents**: 4,043
96
+
97
+ ### Dataset Citation
98
+
99
+ If you utilize this dataset in your research, please cite it as follows:
100
+
101
+ ```
102
+ @dataset{SecAnnual,
103
+ title={SEC Annual Reports (Form 10-K) 1993-2024},
104
+ author={Pleias},
105
+ year={2024},
106
+ description={Collection of SEC annual reports (Form 10-K) for the years 1993 to 2024}
107
+ }
108
+ ```
109
+
110
+ **Note:** This dataset is presented and maintained by Pleias. All rights reserved.