SOLVENT: A Mixed Initiative System for Finding Analogies between Research Papers

Joel Chan·Joseph Chee Chang·Tom Hope
Proc. ACM Hum. Comput. Interact.·2018·85 citations

TLDRSOLVENT is introduced, a mixed-initiative system where humans annotate aspects of research papers that denote their background, purpose, mechanism, and findings, and a computational model constructs a semantic representation from these annotations that can be used to find analogies among the research papers.

SOLVENT: A Mixed Initiative System for Finding Analogies between Research Papers

How do people cite this paper?

(generated 2 months ago)

SOLVENT's annotation scheme for decomposing research papers into background, purpose, mechanism, and findings has been directly adapted for new annotation tasks and extended to finer-grained research aspect classification; its aspect-based document representation has informed work on faceted scientific document retrieval and sentence-level document similarity models; its mixed-initiative approach combining crowdsourcing with computational modeling has served as a baseline and design reference for knowledge modeling systems and crowd-powered annotation tools; its framing of cross-domain analogy discovery has motivated research on semantic relation classification between documents, analogy-based design, and computational creativity support; and its labeled dataset has been reused to build new analogy search models.

Loading PDF reader...

SOLVENT: A Mixed Initiative System for Finding Analogies between Research Papers

Joel Chan·Joseph Chee Chang·Tom Hope
Proc. ACM Hum. Comput. Interact.·2018·85 citations

TLDRSOLVENT is introduced, a mixed-initiative system where humans annotate aspects of research papers that denote their background, purpose, mechanism, and findings, and a computational model constructs a semantic representation from these annotations that can be used to find analogies among the research papers.

SOLVENT: A Mixed Initiative System for Finding Analogies between Research Papers

How do people cite this paper?

(generated 2 months ago)

SOLVENT's annotation scheme for decomposing research papers into background, purpose, mechanism, and findings has been directly adapted for new annotation tasks and extended to finer-grained research aspect classification; its aspect-based document representation has informed work on faceted scientific document retrieval and sentence-level document similarity models; its mixed-initiative approach combining crowdsourcing with computational modeling has served as a baseline and design reference for knowledge modeling systems and crowd-powered annotation tools; its framing of cross-domain analogy discovery has motivated research on semantic relation classification between documents, analogy-based design, and computational creativity support; and its labeled dataset has been reused to build new analogy search models.

Paper

Loading PDF reader...