Voter turnout has been decreasing, and news organizations need to attract voters' interest in elections and encourage them to vote. However, election coverage is not only constrained by law and fairness but also suffers from a shortage of news producers and individualization. Additionally, it remains unclear how voters evaluate information from election coverage. In this paper, we propose a framework that addresses these issues and bridges the gap between election coverage and voters. We demonstrate that data-driven extraction of "unexpected" reasons for winning elections can address the problem of a shortage of news producers and individualization. Furthermore, we show that these reasons effectively engage voters' interest and influence candidate selection, thereby addressing the issue of how voters evaluate such information. By modeling the process by which voters evaluate this information and validating the model, we provide guidelines for news organizations to make decisions on data-driven information for election coverage. Specifically, to attract voters' interest and assist them in selecting candidates, it is important to consider the relationship between the alignment of voters' knowledge and unexpectedness and to explain the information.
Okura et al. (Sat,) studied this question.