Computer Science at NTNU Trondheim (Norway), Professor
講演題目
Finding your way through Petabytes: Evolution and future
trends in search systems
概要
Information retrieval systems based on full text indexing are
traditionally viewed as specialized technology used for web search
engines and few other applications. However, in the past few years,
and continuing today, they are subject to tremendous developments,
focusing on a whole range of current research topics in computer
science, such as:
* Data storage techniques for efficient access to petabytes of
data
* Relevancy and ranking algorithms for large-scale information
retrieval at a consistent sub-second level, i.e. (nearly)
independent of the type of query and the size of the results
* Result set post-processing, navigation, and query recommendation
systems that provide automatic summaries for millions of records
* Meta data aggregation and intelligent retrieval of non-structured
data, based on context-sensitive natural language processing
* Information extraction and high-speed analysis of data streams
The talk will provide insight into several of these developments,
arranged around the concept of Enterprise Search Platform (ESP), which
is rooted in traditional information retrieval, but now increasingly
being used to drive the most challenging, extreme large-scale
information access systems.
Examples of applications will be given, ranging from e-shopping/yellow
pages to adaptive information warehousing and business intelligence.