This research explores how information spreads in web forums by modeling topic-level diffusion using the SIR (Susceptible, Infective, Recovered) epidemic model. An event-driven SIR model is proposed to account for the impact of news postings on information diffusion. The study evaluates these models with a large dataset from a major company's web forum, finding that the event-SIR model better captures the peaks in user participation on major topics.