Re-FRAME the Meeting Summarization SCOPE: Fact-Based Summarization and Personalization via Questions
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917079927488512 |
|---|---|
| author | Kirstein, Frederic Kumar, Sonu Ruas, Terry Gipp, Bela |
| author_facet | Kirstein, Frederic Kumar, Sonu Ruas, Terry Gipp, Bela |
| contents | Meeting summarization with large language models (LLMs) remains error-prone, often producing outputs with hallucinations, omissions, and irrelevancies. We present FRAME, a modular pipeline that reframes summarization as a semantic enrichment task. FRAME extracts and scores salient facts, organizes them thematically, and uses these to enrich an outline into an abstractive summary. To personalize summaries, we introduce SCOPE, a reason-out-loud protocol that has the model build a reasoning trace by answering nine questions before content selection. For evaluation, we propose P-MESA, a multi-dimensional, reference-free evaluation framework to assess if a summary fits a target reader. P-MESA reliably identifies error instances, achieving >= 89% balanced accuracy against human annotations and strongly aligns with human severity ratings (r >= 0.70). On QMSum and FAME, FRAME reduces hallucination and omission by 2 out of 5 points (measured with MESA), while SCOPE improves knowledge fit and goal alignment over prompt-only baselines. Our findings advocate for rethinking summarization to improve control, faithfulness, and personalization. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_15901 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Re-FRAME the Meeting Summarization SCOPE: Fact-Based Summarization and Personalization via Questions Kirstein, Frederic Kumar, Sonu Ruas, Terry Gipp, Bela Computation and Language Artificial Intelligence Meeting summarization with large language models (LLMs) remains error-prone, often producing outputs with hallucinations, omissions, and irrelevancies. We present FRAME, a modular pipeline that reframes summarization as a semantic enrichment task. FRAME extracts and scores salient facts, organizes them thematically, and uses these to enrich an outline into an abstractive summary. To personalize summaries, we introduce SCOPE, a reason-out-loud protocol that has the model build a reasoning trace by answering nine questions before content selection. For evaluation, we propose P-MESA, a multi-dimensional, reference-free evaluation framework to assess if a summary fits a target reader. P-MESA reliably identifies error instances, achieving >= 89% balanced accuracy against human annotations and strongly aligns with human severity ratings (r >= 0.70). On QMSum and FAME, FRAME reduces hallucination and omission by 2 out of 5 points (measured with MESA), while SCOPE improves knowledge fit and goal alignment over prompt-only baselines. Our findings advocate for rethinking summarization to improve control, faithfulness, and personalization. |
| title | Re-FRAME the Meeting Summarization SCOPE: Fact-Based Summarization and Personalization via Questions |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2509.15901 |