Data-driven Methods of Extracting Text Structure and Information Transfer
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909791288295424 |
|---|---|
| author | Honna, Shinichi Murayama, Taichi Matsui, Akira |
| author_facet | Honna, Shinichi Murayama, Taichi Matsui, Akira |
| contents | The Anna Karenina Principle (AKP) holds that success requires satisfying a small set of essential conditions, whereas failure takes diverse forms. We test AKP, its reverse, and two further patterns described as ordered and noisy across novels, online encyclopedias, research papers, and movies. Texts are represented as sequences of functional blocks, and convergence is assessed in transition order and position. Results show that structural principles vary by medium: novels follow reverse AKP in order, Wikipedia combines AKP with ordered patterns, academic papers display reverse AKP in order but remain noisy in position, and movies diverge by genre. Success therefore depends on structural constraints that are specific to each medium, while failure assumes different shapes across domains. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_12999 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Data-driven Methods of Extracting Text Structure and Information Transfer Honna, Shinichi Murayama, Taichi Matsui, Akira Artificial Intelligence Machine Learning The Anna Karenina Principle (AKP) holds that success requires satisfying a small set of essential conditions, whereas failure takes diverse forms. We test AKP, its reverse, and two further patterns described as ordered and noisy across novels, online encyclopedias, research papers, and movies. Texts are represented as sequences of functional blocks, and convergence is assessed in transition order and position. Results show that structural principles vary by medium: novels follow reverse AKP in order, Wikipedia combines AKP with ordered patterns, academic papers display reverse AKP in order but remain noisy in position, and movies diverge by genre. Success therefore depends on structural constraints that are specific to each medium, while failure assumes different shapes across domains. |
| title | Data-driven Methods of Extracting Text Structure and Information Transfer |
| topic | Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2509.12999 |