دانلود مقاله ترجمه شده بازیابی کارآمد اطلاعات در وب معنایی از طریق طراحی معماری – مجله الزویر

بخشی از ترجمه:

سیستم بازیابی اطلاعات وب(IR) که امروزه شاهد آن هستیم، میتواند فقط اطلاعات را بر مبنای واژگان کلیدی بازیابی کند که یک چنین روشی برای حجم زیادی از داده ها ناکارآمد میباشد . چنین روشی دارای قابلیت های محدودی در بدست آوردن مفاهیمِ مربوط به نیازمندی های کاربر و رابطه ی بین واژگان کلیدی مربوطه میباشد. این محدودیت ها باعث ارائه ی ایده ای تحت عنوان جستجوی مفهومی کاربر شده که شامل مفاهیم و معانی میباشد. در این مطالعه قصد داریم سیستم بازیابی اطلاعات معنایی را برای یک جستجوی وب معنایی ارائه داده و الگوریتمی کارمد را برای بازیابی اطلاعات به صورت کارآمد ارائه دهیم. این معماری یک لیستی از واژگان کلیدی که توسط کاربر فراهم شده است را به همراه یک پرس و جو به عنوان ورودی دریافت کرده و آنرا به یک پرس و جوی معنایی تبدیل میکند. این پروسه ی تبدیل با کمک مفاهیم دامنه مربوط به هستی شناسی های دامنه ی موجود صورت گرفته و روابطی معنایی را در زمان اجرا بین آنها پیدا میکند. اطلاعات مربوط به پرس و جوی معنایی نیز بازیابی شده و بر اساس سطح ارتباط آن و با کمک یک الگوریتم بهبود یافته رتبه بندی میشود. تحلیل کارائی نشان میدهد که سیستم پیشنهادی میتواند میزان صحت و کارآمد بودن این روش را برای بازیابی مستندات وب مربوطه در مقایسه با سیستم های موجود بهبود دهد.

بخشی از مقاله انگلیسی:

abstract The current web IR system retrieves relevant information only based on the keywords which is inadequate for that vast amount of data. It provides limited capabilities to capture the concepts of the user needs and the relation between the keywords. These limitations lead to the idea of the user conceptual search which includes concepts and meanings. This study deals with the Semantic Based Information Retrieval System for a semantic web search and presented with an improved algorithm to retrieve the information in a more efficient way. This architecture takes as input a list of plain keywords provided by the user and the query is converted into semantic query. This conversion is carried out with the help of the domain concepts of the preexisting domain ontologies and a third party thesaurus and discover semantic relationship between them in runtime. The relevant information for the semantic query is retrieved and ranked according to the relevancy with the help of an improved algorithm. The performance analysis shows that the proposed system can improve the accuracy and effectiveness for retrieving relevant web documents compared to the existing systems. 2014 Elsevier Ltd. All rights reserved. 1. Introduction 1.1. Motivation The World Wide Web serves as a huge wide distributed global information center for many information services. By now the size of the web is billions of websites and is still growing rapidly. But to get an exact requirement a normal user often spends a lot of time. In order to present the relevant results from this voluminous data to the user, some new methods should be derived to filter the results. The current information technology from the web is mostly based on the keywords. It provides limited capabilities to capture the concept of the user requirement. To solve the limitations of the keyword based search the idea of semantic search is introduced in the field of information retrieval (IR). Information retrieval is the science of searching for documents, information within the documents as well as that of relational database and the World Wide Web. IR also deals with representing, storing and organizing the content. Semantic search has been presented in the IR field since the early eighties (Croft, 1986). The use of ontologies with keyword based search is one of the motivations of the semantic web (SW). The semantic web ‘‘targets to build an extension of the current web in which information is given well-defined meaning, better enabling computers and people to work in cooperation’’ (Berners-Lee, Hendler, & Lassila, 2001). Fig. 1 shows the layers of the semantic web as suggested by Berners-Lee. The bottom layer contains technologies that provide basics for the SW. Uniform resource identifiers (URIs) provide a standard way to refer to entities, while Unicode is a standard for exchanging symbols. The Extensible Markup Language (XML) fixes a notation for describing labelled trees, and XML Schema allows the definition of grammars for valid XML documents. XML documents can refer to different namespaces to make explicit the context (and therefore meaning) of different tags. The middle layer contains technologies to enable building SW applications. Resource Description Framework (RDF) is a framework for creating statements in the form of resources, properties and statements as triples. RDF schema provides a basic vocabulary of RDF. Web Ontology Language describes semantics of RDF statements. SPARQL is RDF Query language. Top layer contain just ideas that should be implemented in order to realize SW. Cryptography, Trust and Proof is to ensure that the http://dx.doi.org/10.1016/j.eswa.2014.07.017 0957-4174/ 2014 Elsevier Ltd. All rights reserved. ⇑ Corresponding author. E-mail addresses: thangarajmku@yahoo.com (M. Thangaraj), sujisekar05@ rediffmail.com (G. Sujatha). Expert Systems with Applications 41 (2014) 8225–8233 Contents lists available at ScienceDirect Expert Systems with Applications journal homepage: www.elsevier.com/locate/eswa Downloaded from http://www.elearnica.ir SW statements are from trusted source. User interface is the final layer that will enable humans to use SW applications. Some of the main goals of the semantic web are Semantic Query Processing and discovering the semantic information available in the unstructured web information with the help of domain ontologies. The IR from the semantic web combines the fast-developing research areas information retrieval, semantic web, and Web content Mining. The various problems associated with the unstructured web pages are identified as follows. (i) Web pages are far complex than that of any traditional document collection. (ii) The web is a highly dynamic information source. (iii) The web serves a broad diversity of user communities. (iv) Only a small portion of the information on the web is truly relevant or useful. These challenges have promoted research into effective and efficient discovery and uses of resources on the internet. There are extensive research activities on the construction and use of semantic web which is nothing but the structure of semantic meaning of the content of the web pages. Web document classification by web mining will help in building the ontology for the semantic web with the automatic extraction of the semantic meaning of web pages. In order to address this issues, the Semantic Based Information Retrieval System (SBIRS) mechanism for SW is proposed. This architecture handles the semantic indexing, extraction, extensions of query and matching of content semantics to achieve the following objectives. (i) Analyze and determine the semantic feature of the content by means of semantic annotation. (ii) Analyze the user’s query and extend it to semantic query using link extraction, ontology and thesaurus. (iii) Match the semantic query with the semantic content using a semantic indexing structure. (iv) Arrange the retrieved results in the order of their relevancy to the query using proposed dynamic ranking algorithm. This architecture eliminates the problems of traditional keyword search and enables the user to retrieve the concept oriented relevant results for any domain. The significance of the framework is to improve search accuracy by understanding searcher intent and the contextual meaning of terms as they appear in the searchable data space. SBIRS also has a few aspects that distinguish it from other related work. Unlike typical search algorithms this framework is based on keywordto-concept mapping with an improved semantic indexing structure and searching technique. The proposed dynamic ranking algorithm presents the results in the order of their relevance for the expanded semantic query. 1.2. Research contributions Contributions of this research fall into the following categories. Clear knowledge of the semantic search, possibilities of semantic enhancements in the IR models. Definition and implementation of a semantic retrieval model with generic domain ontology. Creation of an improved semantic indexing structure. Implementation of a dynamic ranking algorithm Investigation of the feasibility of semantic retrieval in cloud environment. Checking the feasibility of semantic image retrieval. 1.3. Structure of the paper The rest of the paper is organized as follows. An overview of related work is given in Section 2. In Section 3 the working mechanism of the proposed architecture is explained. Section 4 elucidates the retrieval and the ranking algorithms. The performance evaluation is given in Section 5. In Section 6 the main achievements and the tasks that remain is discussed. 2. Related work The unsolved problems of current search engines have led to the development of semantic web search system (Yi Jin & Hongwei Lin., 2008). Conceptual search has been the motivation of a large body of research in the IR field long before the semantic web vision emerged (Jo rvelin, Kekalainen, & Niemi, 2001). ’’SemSearch’’ (Yuangui Lei & Enrico Motta, 2006) is a layered architecture that separates end users from the back-end heterogeneous semantic data repositories. ‘‘SemSearch’’ accepts keywords as input and delivers results which are closely relevant to the user keywords in terms of semantic relations. The SBIRS compliments SemSearch with a ranking algorithm designed specifically for an ontology-based information retrieval model with a semantic indexing structure based on annotation weighing techniques. The inherited relationships between the keywords are analyzed in terms of concepts in ‘‘Ontolook’’ (Li, Wang, & Huang, 2007). From this concepts and relations a concept-relation graph is formed which is used to eliminate the less ranked arcs. It also creates a property-keyword candidate set and sent it to the web page database to get a retrieved result set for the users. The efficiency of this approach is limited by lack of ranking technology. This motivates a relation based page ranking algorithm for semantic web search (Lamberti, Sanna, & Demartini, 2009). The ranking technology is based on the estimate of the probability that keywords/ concepts within an annotated page are linked one with another in a way that is the same to the one in the user’s mind at the time of submitting the query. The probability is measured using a graph-based description of ontology, user query and the annotated page. In these approaches further efforts are requested for future semantic web repositories based on multiple ontologies and better ranking. By building upon a dynamic ontology our model supports multiple domains with semantic dynamic ranking. Fig. 1. Semantic web architecture. 8226 M. Thangaraj, G. Sujatha / Expert Systems with Applications 41 (2014) 8225–8233 An alternate model using ontology is given as an adaptation of vector space model for ontology based information retrieval (Pablo Castells & David Vallet, 2007). This model handles an ontologybased scheme for the semi-automatic annotation of documents and retrieval system. The classical vector space model (Salton and McGill, 1983) is combined with an annotation weighting algorithm and a ranking algorithm. The assumption here is a knowledge base (KB) which has been built and associated to the document base with the help of one or more domain ontologies which describe concepts appearing in the domain text. The documents are then annotated and semantically indexed with the help of the KB. The query is accepted in RDQL form and executed against the knowledge base which returns a list of instances that satisfy the query. This annotation weighting scheme is not considering the different relevance of document fields like title etc. It is addressed in SBIRS by boosting the weight of the query variables in condition on web document tags according to the importance of fields. This work is extended in Miriam Fern‘andez, Vanesa L‘opez, and Pablo Castells (2011), by accepting the input in a formal SPARQL query and later modified in the form of natural language query. This proposal has investigated the practical feasibility of applying semantic search modes in the web environment. Our proposal considers a large scale of heterogeneous web environment. An improved version is discussed in ‘‘The Google Similarity Distance’’ (Rudi Cilibrasi and Paul Vitan‘nyi, 2007) which deals with words and phrases that acquire meaning from the way they are used in society from their relative semantics to other words and phrases. The similarity is based on information distance and kolmogorov complexity. These works are concerned with the similarities between two entities in the ontology. Our approach concerned with the similarity between the query and the web documents. The cluster based approach for information retrieval provides features in terms of reduced size of information provided to the end users. The clusters of items with common semantic and/or other characteristics can guide users in refining their original queries. Users can zoom in on smaller clusters, and then drill down through subgroups (Ramesh Singh & Aman Arora, 2010). Whereas this work is concerned with the query expansion, SBIRS is concerned with starting from query expansion to retrieve ranked results. A Crawler-based indexing and information retrieval system for the semantic web Swoogle (Li Dong et al., 2004) extracts meta data for each discovered document, and computes relations between documents. The ontology rank is computed as a measure of the importance of a semantic web Document. Swoogle is improved by adding user preferences and interests to provide user a set of personalized results. Swoogle is strictly for semantic web documents whereas SBIRS approach converts web documents into semantic web documents. A search engine that uses several mapped RDF ontologies for concept based text indexing is discussed in Jacob Köhler and Michael Specht (2006). For any information retrieval system ranking algorithm is defined with certain metrics. The variety of relevance ranking metrics are discussed and analyzed in Xavier Ochoa (2008). It proposes a set of metrics to estimate the personal, topical and situational relevance dimensions. These metrics are calculated mainly from contextual information and usage and do not require any explicit information from users. Our work moves from the consideration above and relies on the assumption that for providing effective ranking the SBIRS logic makes use of the underlying ontology and of the web page to be ranked in order to compute the corresponding relevance score. A semantic-based approach to content annotation and abstraction for content management is proposed in Hui-Chuan, Chen, and Chen (2009). In this approach a semantic-driven content environment which features a high interoperability of content can be constructed for bridging the semantic gaps for the customer and the content author to increase the efficiency of content management. This work is improved based on the semantics consisting of elements like subject, predicate, and object (Ming-Yen, Chu, & Chen, 2010). In this work a semantic pattern expression capable of representing content semantic features has been designed to represent human semantics with topic-to-topic associations and to replace keywords as the input of the information retrieval system. This architecture contains the core technologies such as Semantic determination and extraction, Semantic Extension, Semantic Pattern and matching. Compared to these approaches, the proposed architecture considers web documents and a much more detailed, densely populated conceptual space in the form of ontology based knowledge base and thesaurus instead of a topic map.