Inktomi Corp. on Wednesday rolled out a new version of its Enterprise Search software as well as a separate edition of the product designed to automate the preparation of content for search engine indexing.
Enterprise Search Version 4.3 is enhanced with tools designed to create a comprehensive search index and improve the relevancy of search results across the enterprise, company officials said.
New features include Content Assistant, an API that enables software developers to index metadata from content management systems and content classification applications with the corresponding source document. This capability improves search results by allowing additional metadata to be injected in a document any time during the indexing process, according to Inktomi officials.
The Quick Links feature enables links to popular documents to be inserted into search results based on keywords, which can help users narrow down searches. Tapping business logic rather than relevance calculations for finding information can improve the efficiency of enterprise search, said Rahul Lahiri, director of product management at Inktomi in Foster City, Calif.
"Search relevance is great but is there is not a way to integrate business logic directly into the relevance algorithms. Quick links allows the integration of business logic into the search result page, which is very helpful for enterprises," he said.
Another enhancement, Binary Large Objects support, allows the search engine to index non-text items from databases such as images, spreadsheets, and PDF files.
To address the problem of preparing enterprise content for search indexing, Inktomi announced its Search Architect Edition. The product combines Inktomi Enterprise Search with a third-party product, Watchfire Enterprise Solution. The combined offering automates the process of refining content before it is searched and allows an administrator to add additional intelligence to the content before it is indexed, Lahiri said.