Content-Path Optimization for TV News Programs

View at publisher Summarise with VIRAT
Authors: Shoma Kaji, Risa Kaneko, Yuta Hagio, Masaru Miyazaki, Siya Bao, Nozomu Togawa
Cited by: 0
References: 2

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This paper formulates the construction of TV news program viewing routes that align with viewer interests while preserving topical continuity as an optimization problem that minimizes semantic distance. Two approaches are proposes: a content-level method that searches for a shortest path under a two-intermediate constraint using pairwise semantic distances computed by a large language model (LLM), and a genre-level method that induces inter-program distances from an LLM-derived inter-genre distance matrix. Evaluations on content sets of size 10 and 40 indicate that the content-level method yields the shortest route between specified endpoints, whereas the genre-level method promotes genre-consistent transitions for start–target pairs with large initial gaps. The results suggest reduced redundant viewing and a balanced trade-off between coverage and coherence for recent events.

Fields of Study & Subjects

Recommender Systems and Techniques Information Retrieval and Search Behavior Video Analysis and Summarization

Publication details

Journal / source
International Conference on Consumer Electronics
Pages
1–3
Publication date
3 February 2026
Language
EN