Generate an ambiguous label for a group of column names, given the table name and database name, such that the label can substitute all the names in the group in natural language questions. The ambiguity should be natural and plausible, making it unclear which specific column the ambiguous label refers to. Return an empty dictionary if there is no semantic correlation between the columns.

Steps

1. Understand Context: Analyze the table and database names to grasp the theme or context.
2. Evaluate Column Names: Review the provided list of column names to identify common themes or overlaps.
3. Construct Ambiguous Label:
   • Identify common words or concepts that the column names might share.
   • Develop a single ambiguous term or phrase that could logically refer to any of the columns.
   • Ensure it is broad enough to fit questions regarding any column in the group plausibly.

Output Format

Provide a list of ambiguous labels, such as a single phrase or a few words. Do not include additional explanations, and keep the format concise.

Examples

Input: Num to generate: 2 Database Name: "UniversityRecords", Table Name: "StudentPerformance", Columns: ["MathScore", "PhysicsScore", "BiologyScore"] Output: ["subject score", "grade"]

Input: Num to generate: 1 Database Name: "HRDatabase", Table Name: "EmployeeDetails", Columns: ["Name", "Surname", "Nickname"] Output: ["personal identifier"]

Notes

• Ensure that the ambiguous label remains a plausible term that might be used in everyday queries or conversations about the topic.
• Avoid overly generic terms unless they are specifically suitable for all elements in the column group.

Num to generate: {num_to_generate} Database Name: {db_name} Table Name: {tbl_name} Columns: {column_group}