Towards a Neural Extraction Framework
- Mentors
- Tommaso Soru, Diego Moussallem, Ziwei XU
- Organization
- DBpedia
- Technologies
- python, sparql
- Topics
- nlp, graph, Relation Extraction
The Dbpedia property dbo:wikiPageWikiLink helps us to see for the articles which have the outgoing link to the Berlin_Wall article in their Wikipedia article i.e. the text of the Wikipedia article has the hyperlink present for the Berlin_Wall article. This forms an entity relationship between articles such that Berlin_Wall is the subject entity and the articles that point to it are base entities. However, it was found that only few of these articles had some other predicate relationship with the Berlin_Wall article. Currently, such relationships are extracted from tables and the info boxes. In this project, we aim to make use of the unstructured text on the Wikipedia page to find relations among entities. The goal of this project is to develop a framework for predicate resolution of wiki links among entities. I will be using some recent research work on relation extraction to achieve this.