Then, how could one separate section 3 material (highly structured data => less interesting) from sections 1 and 2 (high entropy, unstructured => more interesti

Then, how could one separate section 3 material (highly structured data => less interesting) from sections 1 and 2 (high entropy, unstructured => more interesting data) ?