Question

我要从下面的字符串的每一行中提取公司名称，职位名称和职位。有办法吗？因为模式不一致。谢谢。

"Jerry (YC S17) Is Hiring Senior Software Dev, Data Engineer (Toronto/Remote)"

"Iris Automation Is Hiring an Account Executive for B2B Flying Vehicle Software"

"Strikingly (YC W13) is hiring in our Shanghai office"

"BuildZoom (YC W13) is hiring  help make remodeling cheaper"

"EquipmentShare (YC W15) Is Looking for an Experienced React Native Dev"

"Saleswhale (YC S16) AI Assistant Startup Is Hiring Customer Success Managers"

"Streak (YC S11) is profitable, well funded and hiring in Vancouver"

"Tesorio (YC S15) Is Hiring Engineering Managers, Senior Python Engineer"

"Checkr (YC S14) is hiring engineers to build the future of online trust"

"Rescale Is Hiring a Senior DevOps Engineer in San Francisco"

"Tremendous.com is hiring its first engineer"

"Remix is looking for a front-end engineer to help build better public transit"

"Atomwise (YC W15) Is Hiring a Senior Machine Learning Research Scientist in SF"

"Confident Cannabis (YC S15) Is Hiring Engineers"

"WaystoCap (YC W17) is hiring a software engineer in Spain"

"Smarking (YC W15) Is Hiring a Customer Service Manager"

"Sunsama (YC W19) Is Hiring a Senior Full Stack Engineer (RN/GraphQL/Node)"

"Pachyderm Raised $10M and Is Looking for a Senior Full-Stack Engineer"

"Picktrace (YC S15) is hiring a senior Android engineer"

"Segment is hiring engineers to create our developer platform"

"XIX Is Hiring a Senior Front End Engineer"

"Athelas (YC S16) is hiring software engineers"

"Dyneti (YC W19) is hiring software engineers"

"ZeroCater (YC W11) Is Hiring a Principal Engineer in SF: Must Love Food"

"Mux is looking for developers who want to help developers build better video"

"Munich, Germany: Demodesk (YC W19) Is Hiring Software Engineers"

"New Story (YC Nonprofit) Hiring a JavaScript Software Engineer"

"Quit Genius (YC W18) Is Hiring a Product Manager in London"

"Flexport is hiring senior engineers in SF  Come get to know us"

"OneSignal Is Hiring Ruby on Rails and DevOps Engineers in San Mateo"

*************这就是我想要的**************

示例1

“ Jerry（YC S17）正在招聘高级软件开发人员，数据工程师（多伦多/远程）”

公司名称：杰瑞

职位：数据工程师，高级软件开发人员

位置：多伦多/远程

示例2

“ Remix正在寻找前端工程师来帮助建立更好的公共交通”

公司名称： Remix

职位名称：前端工程师

位置：

示例3

“德国慕尼黑：Demodesk（YC W19）正在招聘软件工程师”

公司名称： Demodesk

职位：软件工程师

位置：德国慕尼黑

Answer 1

收集大量此类示例。大约十万个就可以了，但是如果可以得到一百万个样本，那就更好了。
手动拆分它们。如果负担得起的话，将工作分配给实习生/机械土耳其人/等人会更快。
在数据集中训练ML模型。不要忘记随机选择样本进行培训/测试。力争达到90％以上的准确性，但也请尽量不要过度拟合数据。

Answer 2

据我所知，如果模式不一致，则我们无法从任何字符串中提取和获取数据。它只能由具有理解能力的人来完成，或者您需要实施ML。

Answer 3

我们可以轻松使用此模型SPACY，CRF，StanfordNLP和LSTM。训练70％并测试30％的数据。我更喜欢双向LSTM对此最合适。

有没有一种方法可以从下面的每一行字符串中提取“公司名称”，“职务名称”和“职务位置”

3 个答案: