Use of text analysis and topic modelling in policy analysis

In this project we explore the potential use of text anlaysis and topic modelling in assessing and potentially measuring the impact of various methematical modelling efforts in shaping government policy with regard to responding to the COVID-19 pandemic.

We propose to use the Philippines as a case study for this research. Several groups and individuals within the Philippines have proferred predictions on the course of the pandemic in the country. Particularly, these groups or individuals have predicted through their models the possible effects of various non-pharmaceutical interventions on the evolution of the pandemic particularly on the numbers of cases and deaths and their temporal and spatial distribution. Based on these, they have offered recommendations on what should be done further to mitigate these predicted outcomes. On the other hand, the national government has indicated that it will make policies and take action based on scientific evidence and has even commissioned a team of scientists to specifically perform mathenmatical modelling to generate this scientific evidence to guide its policy making.

We propose to quantify the impact of the best available scientific evidence with regard to the COVID-19 pandemic on the policies formulated and enacted by the Philippine government. Specifically, we will measure the degree of fit between the themes/topics focused on by the government’s policies with the themes/topics of the recommendations based on the scientific evidence using text analysis and topic modelling.