Research On Encrypted Data Retrieval Scheme Based On Conceptual Semantics In Cloud Environment

The explosion of information has made it difficult for users to afford the huge cost of local storage,as a result,an increasing number of move their data to the cloud servers.However,the outsourcing of data also enables cloud servers to obtain sensitive data from users,which can bring serious security problem of privacy disclosure.In order to prevent user data from being leaked,sensitive data must be encrypted and then uploaded to the cloud server.However,the complex encrypted technology invalidates the traditional plaintext retrieval technology,therefore,it turns that secure retrieval of encrypted cloud data has become a hot spot recently.Most of the traditional searchable encryption technologies use keywords as the characteristics of documents.Although in recent years,the keyword-based ciphertext retrieval scheme has shown its trend of improvement,the connection gap between keywords and the full-text content of documents makes the retrieval accuracy hardly to meet the needs of users.To reduce the defects above,it is important to realize efficient and reliable encrypted semantic retrieval on the premise of ensuring the data privacy of users.Based on deep research of the semantic searchable encryption retrieval scheme,the main study contents are proposed as follows:The first is the Efficient Content-aware Search based on Semantic Concept over Encrypted Outsourced Data.The scheme obtains the most important topic sentences by calculating document's word embedding,and in the text featuring,it converts the topic sentences into conceptual graph models.In addition,as the matching calculation of semantic graphs is an NPhard problem,a matching algorithm is designed to vectorize the conceptual graphs.Moreover,in order to make the return results more specific,we consider the relationship between concept attributes for the first time,and design the semantic extension of key concept attributes.Finally,the experimental comparison and performance evaluation under the real data set are also advanced points of the program.The second is the Semantic Contextual Search based on Conceptual Graphs over Encrypted Cloud.In this scheme,the relationship between the semantic information of the context and the conceptual graphs is taken into consideration,and a context-based conceptual graph semantic searchable encryption scheme is proposed.The scheme constructs a conceptual attribute weighting algorithm to measure the interest of each concept attribute in the user's mind by querying the words' grammatical relationship.,and balance the functionality and efficiency by extending the semantics of the central concept attributes in the query.Additionally,according to the threat model,this paper proposes two complete encryption retrieval schemes,comparing the scheme with the actual dataset.As a result,it shows the high precision of the scheme.