Abstract: Based on 120 policies related to reading promotion of Northern American public libraries, this
article uses Python and other tools to analyze these policies. Through the high-frequency words statistics,
the co-word matrix and the dissimilar matrix are constructed. Then cluster analysis and multi-dimensional
scaling analysis are carried out. The key contents of reading promotion policy texts of these public libraries
are extracted. The result shows that there are four important elements in these reading promotion policy
texts, including the self- development, outreach missions, the rights and the needs, and the goals of reading
promotion.

Key words: Reading promotion, North American public libraries, Library policy