In 2020, President Tsai Ing-wen gave a speech for re-election. We tried to find things that this government really cares about through the speech text. We also collected the inaugural speeches of every president from 1996, when people can vote for the president. We discovered the changes in relationship with Taiwan and China, and no matter what era, democracy was hot topic.
We could hear what this government wanted to convey to the people from the speech, but from the comparison of speeches over the years, we could find out the order of priority of the previous governments. For example, President Tsai Ing-wen strived for reform during her first term. When she was re-elected, although “change” was still an important word, its importance declined.
We get the content about the president speeches by Python script. And make the segments by Jieba, a Python library for CJK segmentation. After that, we have the term frequency analytics by TF/IDF to identify the keywords of each comment.
What was the hardest part of this project?
Collecting historical data and finding out the changes of the times from various texts are the difficulties of this project.
What can others learn from this project?
We can understand the true meaning behind the words of politicians from the historical changes in the selection of topics for public speaking.