the51job_com
51job.com中文招聘数据集，包含公司信息、职位招聘、福利标签及公司图片来源与归档文件元数据，主要反映2020年采集的招聘网站内容、时间信息及其关联关系。 recruitment employment company-information
README
This database contains a historical snapshot of job and company listings collected from 51job.com, primarily from December 2020. It combines company profiles, job postings, welfare labels, and company image assets.
Scope and organization
The company data describes employers, including names, profile text, business type, location, company size, address details, source URLs, and source-specific counters. The sample includes 贝壳找房（深圳）科技有限公司, a Shenzhen-based real-estate services company. Company records may also include fields that are not listed in the compact schema summary, so source values should be retained rather than assumed to be complete or uniform.
The job data contains individual vacancies linked to employers through company_id. Listings include titles, locations, salary ranges, hiring quantities, posting times, education and experience requirements, tags, addresses, and HTML-formatted descriptions. The sample posting is a Shenzhen real-estate sales role and includes detailed compensation, responsibilities, benefits, and requirements.
Salary values are stored as strings with separate lower and upper bounds and a unit, such as 2–2.5 万/月. These values should not be treated as numeric amounts without interpreting the associated unit. Hiring quantities and some other structured values are also represented as strings. Descriptions may contain embedded HTML, while some fields are empty when the source did not provide a value.
Supporting data
The welfare table stores benefit labels such as 五险两金. It appears to function as a reusable welfare vocabulary; no explicit relationship to a job or company is shown in the supplied schema.
Company imagery is represented in two stages. company_image_origin preserves the original image URL and its associated company, while company_image records metadata for the downloaded asset, including content type, content hash, size, local storage path, title, and a reference to the origin record. The sample image URL uses a protocol-relative 51job CDN address in the origin table and an HTTPS URL in the downloaded-asset table.
Relationships and source identifiers
Jobs reference companies through job.company_id. Company images reference their source URL records through company_image.company_image_origin_id, and the origin records reference companies through company_image_origin.company_id. No direct welfare relationship is present in the supplied schema.
The _yf_slug values preserve source-specific URL or path identifiers. Company values are shown as absolute URLs such as https://jobs.51job.com/all/co5118968.html, whereas job values may be relative paths such as shenzhen-lhq/127410188.html. These identifiers are useful for tracing records back to their source and should not be assumed to be stable canonical keys.
Source and quality notes
Records include collection timestamps in ts, so this should be treated as an archival dataset rather than a current view of 51job.com. The samples were collected on dates in December 2020, and listings, company information, URLs, and image assets may no longer reflect the live site.
Text is primarily in Simplified Chinese, and source content may contain escaped Unicode, HTML markup, URL-encoded terms, or source formatting artifacts. Empty strings, missing values, counters such as _total and _total_worked, and other source-specific fields should be preserved as-is unless a separate normalization process is applied.
Because the data combines crawled pages with downloaded media, timestamps and asset metadata may be recorded at different times. Image hashes and local paths describe the captured files, not necessarily the current contents of the original CDN URLs.
Database Stats:
Database APIs:
-
/company_job_in_china_1/ Companies & Jobs In China 1
Main: /main/list/