linkanalysis is a fundamental task for graph analytics, as it enables the identification of important nodes and patterns in the graph. link analysis algorithms typically require traversing the graph and accessing the...
详细信息
ISBN:
(纸本)9798400708435
linkanalysis is a fundamental task for graph analytics, as it enables the identification of important nodes and patterns in the graph. link analysis algorithms typically require traversing the graph and accessing the links of each node. However, for graphs with a skewed degree distribution, the computing efficiency of linkanalysis is severely constrained due to irregular connectivity, which results in randomized memory accesses and high cache miss ratio. In this paper, we conduct a thorough investigation into skewed graphs' structural characteristics, focusing on their interaction with the computational patterns observed in link analysis algorithms and the underlying hardware architecture. Generalizing the understanding, we develop a novel framework named Mixen. Mixen applies a lightweight filtering procedure to enhance graph locality and reschedule computations. Different graph components are selectively processed under distinctive paradigms. Thereby, Mixen allows for efficient graph traversal and performance optimization. We evaluate Mixen on modern multicore systems, comparing its performance to state-of-the-art frameworks across various graphs and algorithms. The results indicate that Mixen significantly outperforms its counterparts. Over the best-performing alternative, Mixen achieves a 3.42x speedup in execution time. Furthermore, we explore the design space of Mixen, examining its trade-offs and the correlation with cache-memory dynamics.
In the present paper, we consider the problematic of efficiently generating ranked results in the XML IR context, by incorporating the link source of evidence. Despite of their popularity in the Web, only few research...
详细信息
In the present paper, we consider the problematic of efficiently generating ranked results in the XML IR context, by incorporating the link source of evidence. Despite of their popularity in the Web, only few research have exploited links to handle XML IR tasks. In contrast, we propose a new query-dependent linkanalysis approach based on a spreading-activation process that propagates relevance score through the two types of XML links, hierarchical and navigational, to compute a link score for each retrieved XML element. This propagation process depends on two features: the distance between elements and the type of the links separating these elements. The assigned link score is then combined with the content-based score to compute a new score used to re-rank the initial returned list of XML elements. We conducted a series of experiments based on INEX 2007 and 2009 test collections.* Evaluation showed significant improvement compared to baseline runs and previous obtained results.(dagger) These evaluation tests were followed by cross-validation test, which confirmed the robustness of our approach.
This study used data mining techniques to analyze the course preferences and course completion rates of enrollees in extension education courses at a university in Taiwan. First, extension courses were classified into...
详细信息
This study used data mining techniques to analyze the course preferences and course completion rates of enrollees in extension education courses at a university in Taiwan. First, extension courses were classified into five broad groups. Records of enrollees in extension courses from 2000-5 were then analyzed by three data mining algorithms: Decision Tree, linkanalysis, and Decision Forest. Decision tree was used to find enrollee course preferences, linkanalysis found the correlation between course category and enrollee profession, and Decision Forest found the probability of enrollees completing preferred courses. Results will be used as a reference for curriculum development in the extension program. (c) 2006 Elsevier Ltd. All rights reserved.
Web linkanalysis has proven to be a significant enhancement for quality based web search. Most existing links can be classified into two categories: intra-type links (e.g., web hyperlinks), which represent the relati...
详细信息
ISBN:
(纸本)9781581138443
Web linkanalysis has proven to be a significant enhancement for quality based web search. Most existing links can be classified into two categories: intra-type links (e.g., web hyperlinks), which represent the relationship of data objects within a homogeneous data type (web pages), and inter-type links (e.g., user browsing log) which represent the relationship of data objects across different data types (users and web pages). Unfortunately, most linkanalysis research only considers one type of link. In this paper, we propose a unified linkanalysis framework, called "link fusion", which considers both the inter- and intra- type link structure among multiple-type inter-related data objects and brings order to objects in each data type at the same time. The PageRank and HITS algorithms are shown to be special cases of our unified linkanalysis framework. Experiments on an instantiation of the framework that makes use of the user data and web pages extracted from a proxy log show that our proposed algorithm could improve the search effectiveness over the HITS and DirectHit algorithms by 24.6% and 38.2% respectively.
暂无评论