Keyword query routing
Download as DOC, PDF0 likes553 views
This document introduces a novel method for keyword query routing in linked data sources, aiming to reduce processing costs by directing keywords to only relevant sources. It employs a keyword-element relationship summary and a multilevel scoring mechanism to evaluate routing plans, achieving high precision and recall in experiments. The proposed system includes modules for keyword search, query routing, and handling inter-relationships between keywords and data elements.
Keyword query routing
- 1. KEYWORD QUERY ROUTING ABSTRACT: Keyword search is an intuitive paradigm for searching linked data sources on the web. We propose to route keywords only to relevant sources to reduce the high cost of processing keyword search queries over all sources. We propose a novel method for computing top-k routing plans based on their potentials to contain results for a given keyword query. We employ a keyword-element relationship summary that compactly represents relationships between keywords and the data elements mentioning them. A multilevel scoring mechanism is proposed for computing the relevance of routing plans based on scores at the level of keywords, data elements, element sets, and subgraphs that connect these elements. Experiments carried out using 150 publicly available sources on the web howed that valid plan (precision@1 of 0.92) that are highly relevant (mean reciprocal rank of 0.89) can be computed in 1 second on average on a single PC. Further, we show routing greatly helps to improve the performance of keyword search, without compromising its result quality. Algorithm: Keyword Routing Plan: Given the web graph W =(G,N,E) and a keyword query K, the mapping _ : K- 2G that associates a query with a set of data graphs is called a keyword routing plan RP. A plan RP is considered valid w.r.t. K when the union set of its data graphs contains a result for K. The problem of keyword query routing is to find the top-k keyword routing plans based on their relevance to a query. A relevant plan should correspond to the information need as intended by the user.
- 2. EXISTING SYSTEM: Existing work can be categorized into two main categories: There are schema-based approaches implemented on top of off-the-shelf databases. A keyword query is processed by mapping keywords to elements of the database (called keyword elements). Then, using the schema, valid join sequences are derived, which are then employed to join (“connect”) the computed keyword elements to form so-called candidate networks representing possible results to the keyword query. Schema-agnostic approaches operate directly on the data. Structured results are computed by exploring the underlying data graph. The goal is to find structures in the data called Steiner trees (Steiner graphs in general), which connect keyword elements. PROPOSED SYSTEM: We propose to route keywords only to relevant sources to reduce the high cost of processing keyword search queries over all sources. We propose a novel method for computing top-k routing plans based on their potentials to contain results for a given keyword query. We employ a keyword-element relationship summary that compactly represents relationships between keywords and the data elements mentioning them. A multilevel scoring mechanism is proposed for computing the relevance of routing plans based on scores at the level of keywords, data elements, element sets, and subgraphs that connect these elements. Based on modeling the search space as a multilevel inter-relationship graph, we proposed a summary model that groups keyword and element relationships at the level of sets, and developed a multilevel ranking scheme to incorporate relevance at different dimensions. Proposed System architecture
- 3. Modules: The system is proposed to have the following modules along with functional requirements.
- 4. 1. Keyword Search 2. Keyword Query Routing 3. Multilevel Inter-Relationship 4. Set - Level 1. Keyword Search There are schema-based approaches implemented on top of off-the-shelf databases .A keyword query is processed by mapping keywords to elements of the database (called keyword elements). Then, using the schema, valid join sequences are derived, which are then employed to join (“connect”) the computed keyword elements to form so-called candidate networks representing possible results to the keyword query. Schema-agnostic approaches operate directly on the data. Structured results are computed by exploring the underlying data graph. The goal is to find structures in the data called Steiner trees (Steiner graphs in general), which connect keyword elements. 2. Keyword Query Routing We propose to investigate the problem of keyword query routing for keyword search over large number of structured and Linked Data sources. Routing keywords only to relevant sources can reduce the high cost of searching for structured results that span multiple sources. To the best of our knowledge, the work presented in this paper represents the first attempt to address this problem. A solution to keyword query routing can address these problems by pruning unpromising sources and enabling users to select combinations that more likely contain relevant results. For the routing problem, we do not need to compute results capturing specific elements at the data level, but can focus on the more coarse-grained level of sources. 3. Multilevel Inter-Relationship
- 5. The search space of keyword query routing using a multilevel inter-relationship graph. The inter-relationships between elements at different levels are above Fig. A keyword is mentioned in some entity descriptions at the element level. Entities at the element level are associated with a set-level element via type. A set-level element is contained in a source. There is an edge between two keywords if two elements at the element level mentioning these keywords are connected via a path. We propose a ranking scheme that deals with relevance at many levels. 4. Set Level We extract keywords and relationships from the data. Then, based on the elements and sets of elements in which they occur, we create keyword-element relationships. Precomputing relationships (i.e., paths) between data elements are typically performed for keyword search to improve online performance. These relationships are stored in specialized indexes and retrieved at the time of keyword query processing to accelerate the search for Steiner graphs. For database selection, relationships between keywords are also precomputed. This work neither considers relationships between keywords nor relationships between data elements but between keyword-elements that collectively represent the keywords and the data elements in which they occur. Software Requirements: Technologies : Asp .Net and C#.Net Database : MS-SQL Server 2005/2008 IDE : Visual Studio 2008 Hardware Requirements: Processor : Pentium IV RAM : 1GB
- 6. The search space of keyword query routing using a multilevel inter-relationship graph. The inter-relationships between elements at different levels are above Fig. A keyword is mentioned in some entity descriptions at the element level. Entities at the element level are associated with a set-level element via type. A set-level element is contained in a source. There is an edge between two keywords if two elements at the element level mentioning these keywords are connected via a path. We propose a ranking scheme that deals with relevance at many levels. 4. Set Level We extract keywords and relationships from the data. Then, based on the elements and sets of elements in which they occur, we create keyword-element relationships. Precomputing relationships (i.e., paths) between data elements are typically performed for keyword search to improve online performance. These relationships are stored in specialized indexes and retrieved at the time of keyword query processing to accelerate the search for Steiner graphs. For database selection, relationships between keywords are also precomputed. This work neither considers relationships between keywords nor relationships between data elements but between keyword-elements that collectively represent the keywords and the data elements in which they occur. Software Requirements: Technologies : Asp .Net and C#.Net Database : MS-SQL Server 2005/2008 IDE : Visual Studio 2008 Hardware Requirements: Processor : Pentium IV RAM : 1GB