2021/06/01 by Xinyu Hua, Ashwin Sreevatsa, Hua, Xinyu +3
Computer Science · #Computation and Language (cs.CL) #FOS: Computer and information sciences #Natural Language Processing Techniques #Sentiment Analysis and Opinion Mining #Topic Modeling
paper · pdf · doi:10.48550/arxiv.2106.00791
openalex publication_date 2021/06/01 · openalex created_date 2021/06/22 · openalex updated_date 2026/07/28
We study the task of long-form opinion text generation, which faces at least two distinct challenges. First, existing neural generation models fall short of coherence, thus requiring efficient content planning. Second, diverse types of information are needed to guide the generator to cover both subjective and objective content. To this end, we propose DYPLOC, a generation framework that conducts dynamic planning of content while generating the output based on a novel design of mixed language models. To enrich the generation with diverse content, we further propose to use large pre-trained models to predict relevant concepts and to generate claims. We experiment with two challenging tasks on newly collected datasets: (1) argument generation with Reddit ChangeMyView, and (2) writing articles using New York Times' Opinion section. Automatic evaluation shows that our model significantly outperforms competitive comparisons. Human judges further confirm that our generations are more coherent with richer content.