I have a system that receives user queries regarding HR data for their company. Certain…
- User
I have a system that receives user queries regarding HR data for their company. Certain types of questions require different approaches to solving, so we need to classify the queries. What are some classes of question that could be used?
- Assistant
Sure, here are some classes of questions that could be used to classify HR data queries:
-
Job Title/Salary/Benefits: These queries typically involve questions about the job title, salary, and benefits of specific employees.
-
Hiring/Recruitment: These queries typically involve questions about the hiring process and the qualifications of potential candidates.
-
Benefits and Policies: These queries typically involve questions about the benefits and policies of the company, including health insurance, paid time off, and other benefits.
-
Employee Relations: These queries typically involve questions about the company's employee relations policies, including policies regarding grievances, discipline, and other employee-related issues.
-
Payroll Information: These queries typically involve questions about the payroll information of employees, including hours worked, pay rate, and deductions.
-
Benefits Enrollment: These queries typically involve questions about the benefits enrollment process, including questions about the eligibility requirements, forms to fill out, and deadlines.
-
Employee Reviews: These queries typically involve questions about employee reviews, including ratings, comments, and feedback from other employees.
-
Recruitment Process: These queries typically involve questions about the recruitment process, including questions about the qualifications of potential candidates, the application process, and the selection process.
These are just a few examples of the types of HR data queries that could be classified. The key is to identify the specific questions that are being asked and use that information to classify the queries into the appropriate categories.
-
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.