Submissions:2016/Driving Commercial Products with Wikidata: a case study
- Title
- Driving Commercial Products with Wikidata: a case study
- Theme
- Technology and Knowledge Resources
- Type of submission
- presentation
- Author
- Tim Eddo
- E-mail address
- teddo@ntent.com
- Username
- edit9wconf
- Affiliation
- NTENT -- a provider of relevant, mobile content produced by semantic technologies
- Abstract
Is the Wikidata knowledge base a viable source for a commercial technology company to produce quality, mobile content?
NTENT is a commercial company that delivers relevant, multilingual, content to its users internationally. NTENT derives its content from a blend of semantic, machine learning, and natural language processing technologies that must deal with an ever growing number of data sources. To increase the scale of concepts NTENT maintains in its in-house knowledge base, we've embraced the Wikidata project as a basis to seed our popular knowledge domains. However, loosening control over one's carefully curated knowledge resources presents a number of challenges. First and foremost is dealing with the accuracy and consistency of Wikidata's diversely-authored content.
This presentation is a case study in the challenges, and benefits, of building commercial knowledge products from the large-scale, crowd-sourced, Wikidata knowledge base. Notable topics we discuss include: tolerating dirt within Wikidata such as language label quality and ontological inconsistencies; exploiting the knowledge content of underdeveloped Wikidata items while improving them for future use; aligning the identities of Wikidata items with multiple other sources; fusing knowledge from multiple sources with different licensing permissions; and the community benefits of NTENT's dedicated commercial resources towards the improvement of Wikidata content.
At NTENT, we engineer solutions addressing the challenges of integrating open knowledge sources. Our products benefit from the work of Wikidata contributors across the Web. In return, NTENT dedicates professional, commercial resources toward the continual improvement of Wikidata.
- Length of presentation
- 30 min
- Special schedule requests
- Preferred room size
- 25-50
- Will you attend WikiConference North America if your submission is not accepted?
Yes
Interested attendees
If you are interested in attending this session, please sign with your username below. This will help reviewers to decide which sessions are of high interest. Sign with four tildes. (~~~~).
- Minh Nguyễn 💬 11:08, 1 September 2016 (EDT)
- Add your username here.