Research On Semantic Association Data Model And Its Retrieval Mechanisms

With the increasing amount of data resources available in the web, information is becoming more and more complicated. How to efficiently retrieve useful knowledge from the data is still an unresolved yet difficult problem. Semantic association data model gives a desirable solution for tackling this problem. It mainly includes: 1) semantic association data model: the data semantic representation and the model of data organization structure, and 2) more efficient and intelligent semantic retrieval mechanisms based on semantic association data model.This paper proposes the semantic association data model RSS which supports knowledge ranking. The semantic data not only contains resources but also includes the heterogeneous relationships among them. RSS is a novel framework for enabling ranked semantic data retrieval on semantic web. In this framework, the heterogeneity of relationships is fully exploited to determine the global importance of resources. In addition, the search results can be greatly expanded with entities most semantically related to the query, thus able to provide users with properly ordered results by combining global ranking values and the relevance between the resources and the query. The RSS model which supports inference is very different from traditional keyword-based search methods. Moreover, RSS distinguishes from many current methods of accessing the semantic web data in the sense that it applies novel ranking strategies to prevent returning search results in disorder.An effective IR-style keyword based semantic retrieval model is proposed. Current work on semantic data retrieval mainly falls in the SQL-type approach or the so-called semantic search. However, it is still desirable to support flexible keyword search over semantic web, since ordinary users usually do not understand the underlying semantic data structure and also have been accustomed to traditional keyword search for years. In the proposed model, an answer to a keyword query is a connected subgraph that contains all the query keywords. In addition, the answer is minimal because any proper subgraph can not be an answer to the query. An answer has explicit semantics which indicates semantic relationship and strength among keywords. An approximation algorithm with polynomial time complex- ity is presented to retrieve Top-k answers efficiently.This paper presents an intelligent semantic retrieval strategy to address the problem of fuzziness in semantic data. The proposed strategy supports users favoritism among multiple criteria for interesting search results. Due to the intrinsical complexity of real-world knowledge, intelligent retrieval of semantic data to best capture users search intensions still remains a challenging problem. Usually, the information intension of a user search request involves vagueness or imprecision. Moreover, the user may have personalized interests or preferences. A search request is formulated through tightly combining fuzziness with the user's subjective weighting importance over multiple search properties. A special ranking strategy based on weighed fuzzy query representation is presented. It supports users fuzzy and personalized request, thus effectively capturing users underlying information goals.This paper presents a semantic small world (SSW) based semantic data retrieval strategy. Peer-to-Peer systems are gaining popularity in building large-scale information retrieval/sharing systems. However, efficient organizing information and answering specific queries in such systems still remains a challenging problem. In the model, techniques of latent semantic indexing (LSI) and support vector machine (SVM) are applied for document classification summarization, which denotes the documents proportion on each peer. Then, semantically related peers are clustered to form short-range links. Each peer also maintains a very small collection of long-range links which have very strong interests in some particular topics to increase recall and reduce message traffic.