Predictive Analytics for Book Title Selection: A Big Data-Based Study

Abstract

This study presents an integrated, data-driven framework for evaluating publishing titles. It leverages big data analytics to improve editorial decision-making. The architecture features: (1) publisher-survey-calibrated indicator weights optimized with the Analytic Hierarchy Process (AHP); (2) automated pipelines that organize bibliographic data into multi-dimensional repositories, categorizing by author, genre, and time; (3) knowledge graphs using Neo4j to synthesize complex relationships among authors, books, and publishers; and (4) standardized assessment benchmarks, including a composite author proficiency metric. This metric is derived from commercial viability, productivity, and reader perception, each scored on a 0–10 scale.