Toward A More Generalized Web Search Benchmark Workload Model

Web search service is a vital tool to find information on the web. However, not every piece of information found is relevant or useful. In fact, the relevance and precision of the search results is much more important than the time a search engine spends to find the information. From literature, we realize that there is few open or generalized performance evaluation method for web search service. And performance measurement and comparison is critical in the advance of web search technology. Hence, a more generic benchmark approach is needed to assess the relevance and precision of a web search service. The objective of this thesis is to propose a generic construct based approach to develop a more generalized workload model of web search benchmark, and to build an automated benchmarking environment of performance evaluation. We will implement the research model of this research on the benchmark experiment by integrating web search services provided by a web search engine.