Abstract:Judgment aggregation is a framework to aggregate individual opinions on multiple, logically connected issues into a collective outcome. These opinions are cast by judges, which can be for example referees, experts, advisors or jurors, depending on the application and context. It is open to manipulative attacks such as \textsc{Manipulation} where judges cast their judgments strategically. Previous works have shown that most computational problems corresponding to these manipulative attacks are \NP-hard. This desired computational barrier, however, often relies on formulas that are either of unbounded size or of complex structure.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: in the premise - based judgment aggregation framework, the computational complexity of the Manipulation and Bribery problems when using simple formulas.
Specifically:
1. **Limitations of existing research**:
- Previous research has shown that most computational problems related to manipulation and bribery are NP - hard. However, these results usually rely on formulas of unbounded size or complex structure.
- In practical applications, formulas are usually simple and realistic, such as shorter - length clauses, including positive monotone clauses, Horn - clauses, etc.
2. **Research objectives of the paper**:
- Re - evaluate the computational complexity of various manipulation and bribery problems when formulas are restricted to simple, realistic clauses.
- Study the impact of these simple formulas on computational complexity, especially for some basic manipulation variants, whether it can turn problems that were originally known to be NP - hard into problems that are solvable in polynomial time.
3. **Main contributions**:
- For basic manipulation variants, the author shows that these restrictions make some problems that were originally known to be NP - hard become problems that are solvable in polynomial time.
- Provide a P vs. NP dichotomy, covering a wide class of clause restrictions by showing the close relationship between various variants of manipulation and the Satisfiability problem.
- For Hamming distance - based Manipulation, the author proves that for positive monotone clauses of length 3, the problem is still NP - hard, but for positive monotone clauses of length 2, it becomes solvable in polynomial time.
- For the bribery problem, the author shows that even for positive monotone clauses of length 2, the problem is still NP - hard, but in the case of a fixed budget, the problem becomes solvable in polynomial time.
4. **Application scenarios**:
- The paper explains the practical significance of these problems through a specific example (such as Justine's scenario), especially how to avoid the influence of experts' dishonest behavior or mis - evaluation on the final result in financial support decision - making.
In summary, this paper aims to fill the gaps in existing research, provide a more fine - grained computational complexity analysis of manipulation and bribery problems under simple formulas, and thus help understand the security and reliability of these simple formulas in practical applications.