Pattern-Based Acquisition of Scientific Entities from Scholarly Article Titles

CS-KG: A Large-Scale Knowledge Graph of Research Entities and Claims in Computer Science

“…Specifically, we associated the extracted verbs with the high-level predicates of the previous mapping as well as relevant VerbNet classes. We then manually refined this schema to produce a final set of 39 representative predicates mapped to 464 verbs from the articles 16 . These same predicates were also used to produce the relevant relations in the CS-KG ontology.…”

“…Similarly, Nanopublications 6 [19] allow users to represent scientific facts as knowledge graphs and have recently been used to support "living literature reviews", which can be continuously amended with new findings [46]. A common drawback of these solutions is that they are limited to a relatively low number of articles, either because they rely on human experts to summarize information from the literature [24,22] or because they focus on very specific domains (e.g., computational linguistics [16], intrusion detection [48]).…”

Agriculture Named Entity Recognition—Towards FAIR, Reusable Scholarly Contributions in Agriculture

“…To represent scholarly publications as KGs, from an Information Extraction (IE) perspective, named entity recognition (NER) over scholarly publications becomes a vital task since entities are at the core of KGs. As an IE task, NER over scholarly documents is a long-standing task in the NLP community-the Computer Science domain itself has been addressed over a wide body of works with various knowledge capture objectives [6][7][8][9][10][11][12][13][14][15][16][17][18][19]. However, this well-established research area [20][21][22][23], thus far, has not seen any practical applications in the Agricultural scholarly publications domain.…”

“…The ORKG Agri-NER service is an IE system of seven entity types such as research problems, resources, location of study, etc., which since extracted from paper titles implicitly encapsulate the contributions of scholarly articles. Conceptually, the shared understanding around paper titles is that they are succinct summarizations of the contribution of a work [18]. Thus when looking to formulate a contribution-centric entity extraction objective, the first place to seek out this information is from paper titles.…”

SCICERO: A deep learning and NLP approach for generating scientific knowledge graphs in the computer science domain

“…Current solutions either rely on systems for assisting human experts in formalizing their knowledge [2,5] or on information extraction pipelines [11,12]. The first class of solutions is unable to scale and can only be applied to small domains (e.g., computational linguistics [13], intrusion detection [14]). Information extraction techniques can scale, but typically struggle to produce a high-quality output that can be used in a practical setting.…”