Open Domain Web Keyphrase Extraction Beyond Language Modeling
This work addresses keyphrase extraction for web documents across varied domains, offering a dataset and model with incremental improvements over existing methods.
The paper tackles keyphrase extraction from diverse web documents by introducing OpenKP, a large-scale dataset, and BLING-KPE, a model that uses visual features and weak supervision from search queries to handle domain and content variations. Experimental results show BLING-KPE's effectiveness on OpenKP and improved generalization in zero-shot evaluations on DUC-2001.
This paper studies keyphrase extraction in real-world scenarios where documents are from diverse domains and have variant content quality. We curate and release OpenKP, a large scale open domain keyphrase extraction dataset with near one hundred thousand web documents and expert keyphrase annotations. To handle the variations of domain and content quality, we develop BLING-KPE, a neural keyphrase extraction model that goes beyond language understanding using visual presentations of documents and weak supervision from search queries. Experimental results on OpenKP confirm the effectiveness of BLING-KPE and the contributions of its neural architecture, visual features, and search log weak supervision. Zero-shot evaluations on DUC-2001 demonstrate the improved generalization ability of learning from the open domain data compared to a specific domain.