| eG Monitoring |
|---|
|
Measures reported by FSWebAnalyzerTest The Web Analyzer has two main functions: It analyzes search clickthrough logs and hyperlink structures. Both contribute to better ranked search results. Items that show many clicks in the search clickthrough log are popular and therefore receive better rank scores than less-viewed items. Items that are linked to from many other items are also perceived to be more relevant for the user and therefore receive better rank scores. The Web Analyzer improves search relevancy by analyzing the link graph and adding anchor texts and a query independent rank boost based on link cardinality to the items in the index. Anchor texts describe the items they refer to and will improve recall and relevancy when a query term matches the anchor text. Items with many links pointing to them will be ranked higher. The Web Analyzer may scale out to many nodes to reduce the total time that is needed for the analysis. This is done by adding dedicated lookup database components and link processing components that are used during the link analysis. The link processing component receives tasks from the Web Analyzer during link processing. Large scale installations use multiple link processors. The lookup database component represents a key/value lookup server that retrieves the link information generated by the link processing. The item processing looks up the link information for an item using the URL as key. Large scale installations use multiple lookup database components. This test helps administrators analyze the workload of each of the link processing components, and gauge how well each of the lookup/anchor database components will be able to provide link information by reporting the number of sites and URIs that are stored in the database. The measures made by this test are as follows:
|
|||||||||||||||