TAGME Wikification System
(Redirected from TAGME)
A TAGME Wikification System is a Disambiguation to Wikipedia System that is based on Bag-of-Words Mapping.
- Context:
- It solves a TAGME Wikification Task by implementing TAGME Wikification Algorithms.
- It was developed by Ferragina & Scaiella (2010).
- It is composed of the following subsystems:
- Example(s):
- …
- Counter-Example(s):
- See: Natural Language Processing System, WikiText Error Correction System, Co-Reference Resolution System, Part-of-Speech Tagging System, Semantic Role Labeling System, Shallow Parsing System, Text Analysis System, Document to Ontology Interlinking System, Wikimedia, Disambiguation to Wikipedia Task.
References
2010
- (Ferragina & Scaiella, 2010) ⇒ Paolo Ferragina, and Ugo Scaiella. (2010). “TAGME: On-the-fly Annotation of Short Text Fragments (by Wikipedia Entities).” In: Proceedings of the 19th ACM International Conference on Information and knowledge management. ISBN:978-1-4503-0099-5 doi:10.1145/1871437.1871689
- QUOTE: We designed and implemented TAGME, a system that is able to efficiently and judiciously augment a plain-text with pertinent hyperlinks to Wikipedia page (...) a software system that annotates short fragments of text on-the-fly and with high precision by cross-referencing some meaningful text spots (i.e. anchors drawn from Wikipedia) with pertinent concepts (i.e. Wikipedia pages). This is obtained in two main phases, namely anchor disambiguation and anchor pruning. Disambiguation will be based on finding the “best agreement” among the concepts assigned to the anchors detected in the input text, while pruning will evaluate the “significance” of an anchor-sense annotation by considering the relatedness among the assigned senses and their link probability, in order to possibly drop the annotations which result non-pertinent with the topics the input texts talks about.
So the structure of TAGME mimics the one of [1], [2]’s systems and, as done in these papers, it aims at the collective agreement among all anchor annotations. However, unlike prior works (...), TAGME deploys some new scoring functions that evaluate the “significance” of an anchor annotation in a way that is fast and effective in the quality of the final assignment.
- QUOTE: We designed and implemented TAGME, a system that is able to efficiently and judiciously augment a plain-text with pertinent hyperlinks to Wikipedia page (...) a software system that annotates short fragments of text on-the-fly and with high precision by cross-referencing some meaningful text spots (i.e. anchors drawn from Wikipedia) with pertinent concepts (i.e. Wikipedia pages). This is obtained in two main phases, namely anchor disambiguation and anchor pruning. Disambiguation will be based on finding the “best agreement” among the concepts assigned to the anchors detected in the input text, while pruning will evaluate the “significance” of an anchor-sense annotation by considering the relatedness among the assigned senses and their link probability, in order to possibly drop the annotations which result non-pertinent with the topics the input texts talks about.