评估商业数据库的完整性：使用网络抓取技术与官方记录进行比较.pdf

1、WIN2025Evaluating the completeness of business databasesWIN2025Evaluating the completeness of business databasesJosep Domnech Universitat Politcnica de ValnciaEvaluating the completeness of business databasesA comparison with official records using web scraping techniquesXin-Hui Huang,Josep DomnechU

2、niversitat Politcnica de ValnciaWIN2025Evaluating the completeness of business databasesMotivation SABI is widely used for business research It is the local version of the ORBIS database BORME,the official business registry,provides a complementary source.Aim:Compare BORME and SABI to identify missi

3、ng companies.2WIN2025Evaluating the completeness of business databasesResearch objectives Assess how many companies established 20102023 are missing in SABI Examine coverage by year of establishment,dissolution,province,sector,and legal form Uncover potential biases (e.g.,survivorship bias,size bias

4、)3WIN2025Evaluating the completeness of business databases SABI(Bureau van Dijk)Commercial database with financial and registration data Contains around 1.9 million Spanish companies(as of January 2024)Data SourcesWIN2025Evaluating the completeness of business databases BORME(Official Gazette in Spa

5、in)Daily PDF publications by each registry Essential legal information on all newly registered and dissolved companiesData SourcesWIN2025Evaluating the completeness of business databasesMethodology Downloaded and parsed 100k PDF publications from BORME Extracted 9.96 million entries on 3.05 million

6、firms Entries were divided in fields(establishment,dissolution,legal form)Filtered for companies established 20102023 1.30 million firms Matched with SABI records to identify missing entitiesWIN2025Evaluating the completeness of business databasesCoverage Only 38.3%of firms created 20102023 found in