Date of Award
Spring 2024
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Statistics and Data Science
First Advisor
Crawford, Forrest
Abstract
Crowdfunding has become a popular alternative to traditional financing for individuals and organizations seeking to raise funds from a large group of backers. Managers need information about dynamics of their platforms to optimize operations. The Hawkes process has been used to represent self-exciting phenomenon in a variety of domains, but has not been applied to crowdfunding in a way that could help managers understand crowd behaviors. We extend the Hawkes process to capture important properties of crowdfunding platform dynamics, and apply the model to analyze data from two donation-based platforms. A background review of point processes, structured around the three components, item, user and contribution, is given to contextualize the model construction. For each user-item pair, the continuous-time conditional intensity is modeled as the superposition of a self-exciting baseline rate and a mutual excitation by preferential attachment, both depending on prior user engagement, and attenuated by a power law decay of user interest. The model is thus structured around two time-varying features, contribution count and item popularity. Our approach based on item relative popularity is compared with the computational complexity of the double summation in the standard Hawkes process likelihood. Parameters are determined with maximum likelihood estimation from 2,000 items and 164,000 users over several years. We identify a bottleneck in the user contribution pipeline, measure the force of item popularity, and characterize the decline in user interest over time. A contagion effect is introduced to assess the concentration on popular items. Simulations are conducted using the inverse transform method. Goodness-of-fit assessment is achieved by a transformation into a Poisson process with the random time change theorem. This mechanistic model lays the groundwork for enhanced crowdfunding platform monitoring based on evaluation of counterfactual scenarios and formulation of dynamics-aware recommendations.
Recommended Citation
Djorno, Alexandra, "Extension of the Hawkes Process for Modeling Crowdfunding Platform Dynamics" (2024). Yale Graduate School of Arts and Sciences Dissertations. 1293.
https://elischolar.library.yale.edu/gsas_dissertations/1293
