|
|
|
| Research on Enterprise Risk Information Extraction Method Based on Triple Dimensions |
| Liang Na1, Yao Changqing1, Wang Zheng2, Gao Yingfan1, Li Yan1 |
1.Institute of Scientific and Technical Information of China, Beijing 100038
2.National Science Library, Chinese Academy of Sciences, Beijing 100190 |
|
|
|
|
Abstract In recent years, the length of the annual report of enterprises has been increasing. As the main body of the annual report, the content of the three financial statements has remained constant. However, the text content outside the financial statements has become increasingly descriptive. Various supplementary explanations have become a useful addition to understand the current situation of the company’s production and operation. Among them, risk information disclosure has gradually become the focus of scholars due to its forward-looking and decision-making relevance. As such, the question of how to extract valuable content from the large amount of risk information has become a problem worth studying. Therefore, this paper considers the risk information disclosed in the 2016 semi-annual report of all A-share listed companies as the background data. It then proposes a three-dimensional risk information extraction method and extracts the risk information from the risk description text so that the extracted risk phrases are rich. The information content is extracted as much as possible to restore the information to be expressed in the original risk text.
|
|
Received: 22 May 2019
|
|
|
|
1 NordenL,WeberM. Credit line usage, checking account activity, and default risk of bank borrowers[J]. Review of Financial Studies, 2010, 23(10): 3665-3699.
2 AthanasakouV,HussaineyK. Forward-looking performance disclosure and earnings quality[R]. London School of Economics and University of Stirling Working Paper, 2010.
3 AthanasakouV,HussaineyK. The perceived credibility of forward-looking performance disclosures[J]. Accounting and Business Research, 2014, 44(3): 227-259.
4 GulinD,HladikaM,Mi?inM. Disclosure of non-financial information: The case of croatian listed companies[C]// Proceedings of the Conference on Consumer Behavior, Organizational Strategy and Financial Economics. Cham: Springer, 2018, 9: 159-175.
5 BochkayK,LevineC B. Using MD&A to improve earnings forecasts[J]. Journal of Accounting, Auditing & Finance, 2019, 34(3): 458-482.
6 林钟高, 杨雨馨. 风险提示信息与银行信贷决策——基于A股上市公司年报文本信息的研究[J]. 安徽师范大学学报(人文社会科学版), 2017, 45(2): 245-255.
7 孟庆斌, 杨俊华, 鲁冰. 管理层讨论与分析披露的信息含量与股价崩盘风险——基于文本向量化方法的研究[J]. 中国工业经济, 2017(12): 132-150.
8 申心吉. 中国上市公司信息披露质量状况研究——基于深交所信息披露考评的经验证据[J]. 时代金融, 2017(12): 152-154.
9 LiF. Textual analysis of corporate disclosures: A survey of the literature[J]. Journal of Accounting Literature, 2011, 29: 43-65.
10 HanleyK W,HobergG. The information content of IPO prospectuses[J]. Review of Financial Studies, 2010, 23(7): 2821-2864.
11 DeAngelisM D. Uncommon information in firm disclosures[D]. East Lansing: Michigan State University, 2014.
12 YangB. Extending topic models for text analysis of corporate risk disclosures[D]. Singapore: National University of Singapore, 2013.
13 翟文洁, 闫琰, 张博文, 等. 基于混合深度信念网络的多类文本表示与分类方法[J]. 情报工程, 2016, 2(5): 30-40.
14 CampbellJ L,ChenH,DhaliwalD S, et al. The information content of mandatory risk factor disclosures in corporate filings[J]. Review of Accounting Studies, 2014, 19(1): 396-455.
15 周双文. 基于领域本体的创业板公司年报风险信息抽取方法研究[D]. 长沙: 湖南大学, 2013.
16 胡小荣, 姚长青, 高影繁. 基于风险短语自动抽取的上市公司风险识别方法及可视化研究[J]. 情报学报, 2017, 36(7): 663-668.
17 肖浩, 詹雷, 王征. 国外会计文本信息实证研究述评与展望[J]. 外国经济与管理, 2016, 38(9): 93-112.
18 张秋子, 陆伟, 程齐凯, 等. 基于最大熵模型的学术缩写自动识别[J]. 情报工程, 2015, 1(2): 64-72. |
|
|
|
