2017
DOI: 10.1504/ijmso.2017.090778
|View full text |Cite
|
Sign up to set email alerts
|

Linking science: approaches for linking scientific publications across different LOD repositories

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
4
0

Year Published

2018
2018
2024
2024

Publication Types

Select...
3
3
1

Relationship

1
6

Authors

Journals

citations
Cited by 8 publications
(4 citation statements)
references
References 0 publications
0
4
0
Order By: Relevance
“…As seen in the example of figure 1, the title is factorized more than other elements. Based on our previous work [21], as the most determinant combination, we have perceived the combination of all of them by doubling the weight/importance/impact of the title. The title is often most representatives, as authors tend to include the key terms regarding the subject in it.…”
Section: A Searchingmentioning
confidence: 99%
See 1 more Smart Citation
“…As seen in the example of figure 1, the title is factorized more than other elements. Based on our previous work [21], as the most determinant combination, we have perceived the combination of all of them by doubling the weight/importance/impact of the title. The title is often most representatives, as authors tend to include the key terms regarding the subject in it.…”
Section: A Searchingmentioning
confidence: 99%
“…Hence, the model is trained on a corpus of around 12 million words, with a windows size of 5 and 300 dimensions. Based on our previous work [21], building a model on top of a specific domain-related datasets, the Word2Vec gives closely related domain correlations of terms; whereas the existing pre-trained models such as Google News provides more general context to a particular term. Figure 4 shows an example of Word2Vec implementation for suggesting top five most similar terms, considering the term "globalization".…”
Section: B Selective Searchmentioning
confidence: 99%
“…They also discussed various ways to enhance the interoperability of RIs, including semantic contextualization, enrichment, mapping and bridging. Hajra et al [15] presented a way to enhance scholarly communication by linking data from different repositories. They considered bibliographic Linked Open Data (LOD) repositories to compute the semantic similarity between two resources.…”
Section: Introductionmentioning
confidence: 99%
“…They have tested these strategies for multiple tasks, such as classification and regression, by using two datasets Wikidata and DBpedia, whereas they have used the GloVe model [195] for creating RDF embeddings by exploiting global patterns. The authors in [111] used several bibliographic RDF datasets and word2vec for enriching the data of scientific publications with information from multiple data sources, while in [123] the authors exploited enriched ontology structures for producing RDF embeddings which were used for the task of Entity Linking. In [177], the authors combined embeddings from DBpedia and social network datasets for performing link prediction, whereas in [72] Wikipedia knowledge graph was exploited for finding the most similar entities to a given one for a specific time period.…”
Section: Related Workmentioning
confidence: 99%