LLMs as Architects and Critics for Multi-Source Opinion Summarization

benchmark dataset for evaluating multi-source opinion summaries across 7 key dimensions: fluency, coherence, relevance, faithfulness, aspect coverage, sentiment consistency, specificity. Our results demonstrate that M-OS significantly enhances user engagement, as evidenced by a user study in which, on average, 87% of participants preferredM-OS over opinion summaries. Our experiments demonstrate that factually enriched summaries enhance user engagement. Notably, M-OS-PROMPTS exhibit stronger alignment with human judgment, achieving an average Spearman correlation of ρ = 0.74, which surpasses the performance of previous methodologies.

Opinion Summary

The racquet is highly praised for its lightweight design and powerful performance,

particularly in smashes, making it ideal for both offensive and

defensive play. Many users recommend it for intermediate to advanced

players, especially those on a budget. The head-heavy design boosts power

while ensuring easy movement, with durable military-grade carbon construction.

Some users note minor drawbacks, such as a smaller grip size,

which can easily be adjusted with external grips. However, one customer

reported receiving a damaged racquet with no replacement policy. The

racquet offers excellent performance, durability, and a lightweight design.

Multi-Source Opinion Summary

The Li-Ning G-Force Racquet is designed for advanced players prioritizing

speed and precision. Crafted from Japanese Ultra-Carbon Graphite, it

ensures exceptional strength and rigidity, preventing frame deformation.

The UHB Shaft designed using player data, optimizes smash performance,

while the oval 53.5 sq/in head size and G2 grip (9.52 cm) provide control

and maneuverability. The Aero Tec Beam System reduces air resistance

while preserving frame integrity, and the Dynamic Optimum Frame design

boosts bounce strength for powerful smashes. Weighing only 79g with

a head-heavy balance, it excels in both offensive and defensive play. At

26.7 inches tall with a 7 mm beam width, it delivers top-tier performance.

With an average rating of 4.2, users praise its durability, though some

recommend external grips. Despite a rare issue with a broken racquet, it

remains a top choice for players seeking precision and power.

M-OS enriches summaries by incorporating technical specifications and product descriptions, enabling precise product comparisons - a key advantage over review-only approaches that often lack detailed attributes. M-OS addresses decision fatigue and information overload by synthesizing diverse product data to deliver comprehensive, relevant summaries