Computational-Content-Analysis-2018 / 12-Jan-Quantitative-Analysis-of-Culture-Using-Millions-of-Digitized-Books

Michel, Jean-Baptiste et al. 2010. “Quantitative Analysis of Culture Using Millions of Digitized Books.” Science express, December 16.
https://github.com/Computational-Content-Analysis-2018
0 stars 2 forks source link

Culture in Millions of Digitized Books #14

Open sunnyJy opened 6 years ago

sunnyJy commented 6 years ago

(1) Enriched meaning behind words: Figure 1 (A) plot the "apple" frequency. What if new meanings were created in addition to the original meaning? At the very beginning, "apple" is a fruit name. It may appear in history books (e.g. how apples were discovered and spread to different places), biology/farming books (e.g. how to cultivate apples), medical books (e.g. an apple a day keeps doctor away), etc. However, when Newton came, "apple" were mentioned as the object hitting his head, which may appear on physics books. In 2000s, Steve Jobs developed Apple company, people are in favor of Apple products. Nowadays, Apple products are everywhere, Newton's story is inscribed in all the scientific books. How to take this into account while doing analysis?

(2) Question about fame: Fame is measured by frequency of people's names. It does not classify/ define whether the “fame” is an indicator of success or notoriety. Do you think it's necessary to take this into account?

(3) Data Representativeness: Topics books have to be as wide as possible because words frequencies may correlate with the book topics. And the possibility of being digitized may correlated with the book topics as well.

(4) From the impact of censorship on history (9%) literature (27%), politics (60%), philosophy (76%), and art (56%), I could roughly know the level of impact of history, literature, philosophy and art on the politics.

(5) What's the definition of "half-life"?

-- suunyJY 2018.1.11