Disambiguation to Wikipedia: A Language and Domain Independent Approach

Nguyen, Truc-Vien T.

doi:10.1007/978-3-642-45068-6_27

Cited by 1 publication

(4 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We decided to develop our own systems for these three tasks. Specifically, to identify private data in Facebook statuses, we made use of our previous works presented in [23,24] and of a new method presented in [22]. In the former, we developed a system which recognizes named entities in text [23] and associate them with Wikipedia entries [24].…”

Section: Semantic Analysismentioning

confidence: 99%

“…In the former, we developed a system which recognizes named entities in text [23] and associate them with Wikipedia entries [24]. Recently we proposed the system in [22], which solves the D2W problem in a language-independent manner. Thus, our D2W system can be applied in any language available in Wikipedia, as opposed to one single language observed in all previous works [5,20,21,27].…”

Section: Semantic Analysismentioning

confidence: 99%

“…Since Wikipedia mainly contains popular names of people, organizations, locations, if a name can be linked to Wikipedia, it is very likely that it is a popular name. We use the NER system of [23] and the D2W system developed in [22].…”

Section: Detection Of Private Datamentioning

confidence: 99%

“…In the Policy Ontology for connecting a policy (which is an individual) to its activation condition (a class) we define a specific property: hasPolicyActivation: Policy → Object. This is possible in OWL 2 thanks to punning 22 Class ↔ Individual, which allows to use the same term for both a class and an individual.…”

Section: The Model Of Policiesmentioning

confidence: 99%

See 3 more Smart Citations

Automatic policy enforcement on semantic social data

Nguyen

Fornara²,

Marfia³

2015

MGS

Self Cite

View full text Add to dashboard Cite

Web-based data collection of non-reactive data is becoming increasingly important for many social science fields. Being able to introduce and automatically enforce policies that regulate the collection and the use of those data is crucial for taking into account the privacy and confidentiality wishes of data providers. Those policies are currently expressed in natural language or in a language lacking a formal semantics, therefore, in order to comply with them researchers need to apply them manually. It is therefore clear that realizing new technologies for automatically supporting the enforcement and monitoring of various type of policies in Web-based data collection is an important open issue. In this paper, we propose an application independent formal model of policies that can be automatically enforced and whose specification allows to take into account the meaning of the managed data. Given that for enforcing the policies on the collected data, their semantics is crucial, in this paper we propose an OWL 2 Social Network Ontology to capture the nature of social network data collected from the Web. We also propose techniques for semantic enrichment of the data, which are then exploited in the specification and enforcement of the formalized policies.

show abstract

Section: Semantic Analysismentioning

confidence: 99%