### ROLE
You are an expert system specializing in knowledge graph linguistics. Your task is to transform structured data (subject-relation-entity triples) into natural, grammatically correct English questions.

### GOAL
Given three lists of aliases (for a subject, a relation, and an answer entity), you must generate a high-quality question-answer pair based on this data.

### INSTRUCTIONS
Follow these steps precisely:

**Step 1: Select the Best Aliases**
Your first task is to choose the single best alias from each list for human-readable output.
- **From `Question subject aliase`:** Select the most common, canonical, and easily readable name. Avoid abbreviations, technical IDs, or non-standard capitalization (e.g., choose "Indiana University" over "iu.edu" or "indiana university").
- **From `Question relation aliase`:** Select the verb or phrase that conveys as much information as possible in a concise manner. (e.g., "borders" is often better than "shares border with"; "is a member of" is better than just "member of" for forming a question).
- **From `Question answer aliase`:** Select the most common and canonical name for the answer entity (e.g., choose "France" over "FR" or "République française").

**Step 2: Formulate the Question**
This is the most critical step. Create a natural, grammatically perfect question in English.
- **Use the chosen subject alias** as the main subject of the question.
- **Infer the semantic type of the answer entity** to choose the correct question word (Who, What, Which, Where, etc.).
    - If the answer is a person (`human`), use "Who...".
    - If the answer is a country, use "Which country...".
    - If the answer is an organization, use "Which organization..." or "What organization...".
- **Crucially, DO NOT simply force the relation alias into a template.**
    - **Bad Example (from relation 'is a'):** "Who is a Lalit Kumar Goel?"
    - **Good Example:** "Who is Lalit Kumar Goel?" or "What is Lalit Kumar Goel?"
    - **Bad Example (from relation 'borders'):** "The Kingdom of the Netherlands shares border with what?"
    - **Good Example:** "Which country borders the Kingdom of the Netherlands?"
    - **Bad Example (from relation 'member of'):** "Member of which association is Indiana University?"
    - **Good Example:** "Which association is Indiana University a member of?"

**Step 3: Format the Output**
Combine your results into a single line, using `[END]` as the dividing token. The sequence MUST be:
`[Chosen Subject Alias] [END] [Chosen Relation Alias] [END] [Generated English Question] [END] [Chosen Answer Alias]`

### EXAMPLES
These examples demonstrate the application of the above rules.

**Example 1:**
*   **Input:**
    *   Question subject aliase: ['Reino Hulandes', 'kingdom of the netherlands', 'The Kingdom of the Netherlands', 'Koninkrijk der Nederlanden']
    *   Question relation aliase: ['shares border with', 'borders', 'bordered by', 'adjacent to']
    *   Question answer aliase: ['iso 3166-1:fr', 'FR', 'France', 'république française']
*   **Correct Output:**
    Kingdom of the Netherlands [END] borders [END] Which country borders the Kingdom of the Netherlands? [END] France

**Example 2:**
*   **Input:**
    *   Question subject aliase: ['lalit kumar goel', 'Lalit Kumar Goel']
    *   Question relation aliase: ['instance of', 'is a', 'is an', 'has type', 'rdf:type']
    *   Question answer aliase: ['Huamn', 'Humans', 'human beings', 'human', 'person']
*   **Correct Output:**
    Lalit Kumar Goel [END] is a [END] Who is Lalit Kumar Goel? [END] human

**Example 3:**
*   **Input:**
    *   Question subject aliase: ['Indiana U', 'Indiana University System', 'indiana.edu', 'Indiana University']
    *   Question relation aliase: ['member of']
    *   Question answer aliase: ['Association of research libraries', 'Association of Research Libraries', 'ARL']
*   **Correct Output:**
    Indiana University [END] member of [END] Which association is Indiana University a member of? [END] Association of Research Libraries

### DATA
Question subject aliase: %s
Question relation aliase: %s
Question answer aliase: %s
