Verifiable Keyword-Based Semantic Similarity Search on Social Data Outsourcing

Abstract

In the data-driven economy era, social data have tremendous business and potential values. Obtaining authentic social data is the first step in mining the business value of social data. In this paper, we consider an emerging social data outsourcing paradigm. Therein, different online social network (OSN) operators outsource their social data to a third-party social data provider (SDP), who resells them to data consumers who can be any individual or entities. However, a dishonest SDP may return untrusted query results to data consumers through various activities, such as adding fake data and deleting/modifying correct data. To deal with these dishonesties, we propose a basic scheme and an enhanced scheme to allow data consumers to verify the correctness and completeness of their received social data from the SDP. Data consumers in the basic scheme utilize the public APIs to collect the sampled social data and compare them with their received social data. This scheme is a probabilistic verification method as the data consumers only having a tiny proportion of the social data. To permit data consumers to verify the query results trustworthiness deterministically, we proposed an enhanced scheme, in which the OSN operator generates some cryptographic auxiliary information. The SDP can construct a verification object for the data consumer based on these information. Extensive experiments ran on a real Twitter dataset confirm that our schemes are effective and efficient.

Keywords

• Social data outsourcing
• verifiable keyword-based search
• semantic similarity measurements