CS 388 Momo Hirose Pitches
by on with No Comments

Idea #1 

My research question: 

What algorithms can be used to simulate policy making?

Public policy should obviously be for all citizens, but people often feel that their voices and real needs are not reflected in policy. In my home country of Japan, I have witnessed many voices on Twitter (now called X) about the policies they want from the government. I would like to create software that can use government data to simulate and predict what the impact would be if the government actually implemented those policies. Such a tool would help in better evidence-based policy making.

Idea #2

My research question: 

How can we combine speech recognition, AI, and data extraction to automatically and instantly display the data mentioned in public discussions?

In Japan, when public policies are made, they have to be discussed in the Council or the Diet (Japanese national congress) in order to be approved. However, these discussions are often not easy to understand for the citizens for the following two reasons. 

  1. Politicians tend to criticise each other for trivial words or actions, so it is hard to understand the points of what is being discussed.
  2. What we, the citizens, see on television is only politicians arguing. Without visual aids it is even harder to follow the discussions. 

If the data (graphs and numbers) are automatically displayed when politicians refer to data by recognising the voice, policy-making will be more evidence-based and discussions will be easier for citizens to understand.

  • An example of a specific issue I would like to look at is Japan’s declining birth rate – this is often discussed in the National Diet, but because it is such a broad topic, the discussions sometimes get lost without data.
    • https://www.shugiintv.go.jp/jp/
    • * This link above is the website where you can access all the archives of live broadcasts of the National Diet.
    • * The videos cannot be downloaded (only available for watching in the web browser) 
    • https://www.shugiin.go.jp/internet/itdb_shitsumon.nsf/html/shitsumon/menu_m.htm
      • This government official website shows text data of the questions and answers that were discussed in the National Diet. They are listed by topics and available in PDF and HTML format. (* Only in Japanese, but I can translate for the purpose of this research) 

Idea #3  Computational Modeling of National Budget

My research question: 

What algorithms can be used to better allocate the national budget?

What I realized while working with the government agencies in Japan last summer was that the policy-making process focuses heavily on the rationale of policies, and after they are approved, the evaluation is not as important as it should be. This could lead to a situation where the national budget continues to be allocated to policies that are not effective. If we can create a system to quantify the impact of each policy and also get comments from citizens, and use an algorithm to allocate the national budget, I think we could have a better cycle of public policy.

  • Since examining the entire national budget could be huge, here’s a specific topic I’m interested in investigating: 
    • The budget for developing the startup ecosystem in Japan
      • How is it currently allocated to each policy?
      • How effective are these policies?
      • How should the budget be allocated in the future, using computer modelling?
      • * Data and information on the development of the startup ecosystem and its budget can be found at the Japanese Ministry of Economy, Trade and Industry.