Xiannong Meng

A Study of Color Histogram Based Image Retrieval

This paper describes a project that implements and tests a simple color histogram based search and retrieve algorithm for images. The study finds the technique to be effective as shown by analysis using the RankPower measurement. The testing also highlights the weaknesses and strengths of the model, concluding that the technique would have to be augmented and modified in order for practical use.

WebSail: from on-line learning to Web search

In this paper we report our research on building WebSail, an intelligent web search engine that is able to perform real-time adaptive learning. WebSail learns from the users relevance feedback, so that it is able to speed up its search process and to enhance its search performance. We design an efficient adaptive learning algorithm TW2 to search for web documents. WebSail employs TW2 together with an internal index database and a real-time meta-searcher to perform real-time adaptive learning to find desired documents with as little relevance feedback from the user as possible. The architecture and performance of WebSail are also discussed.

A Comparative Study of Performance Measures for Information Retrieval Systems

Traditional performance measures of information retrieval systems include precision and recall and their variants. While these measures work well in closed-laboratory environments, they are not suitable for practical IR systems such as Web search systems. Many single-value measures were proposed to improve over the precision-recall measure, such as expected search length (ESL), average search length (ASL) and RankPower. We compare in this paper the measures of ESL, ASL, and RankPower applied to a set of real Web retrieval data. The results demonstrate that RankPower indeed is a feasible, effective, and easy-to-use single-value measure for performance of practical IR systems such as Web search engines

An empirical user rating of popular search engines using RankPower

We present in this paper the effectiveness and the feasibility of measuring search engine efficiency using RankPower, a single value measurement. RankPower rates the search engines by computing the number and the rank of relevant URLs among the returned results. This is different from traditional performance measures of information retrieval systems using the precision and recall rates. The RankPower is much more easier to compute and much more intuitive for ordinary Web search users. The conclusions are supported by a set of empirical results collected from the experiments that the authors designed and carried out.

A multiplicative gradient descent search algorithm for user preference retrieval and its application to Web search

The gradient descent procedure of Wong et al. (1988) for user preference retrieval is based on linear additions of documents judged by the user. In contrast we design in this paper a multiplicative gradient descent search algorithm MG that uses a multiplicative query expansion strategy to adaptively improve the query vector. Our work generalizes the work of Wong et al. in the following two aspects: various updating functions may be used in our algorithm; and multiplicative updating for a weight is dependent on the value of the corresponding index term, which is more realistic and applicable to real-valued vector space. The algorithm MG boosts the usefulness of an index term exponentially, while the algorithm of Wong et al. does so linearly. We report a working prototype of the Web search project MAGRADS (Multiplicative Adaptive Gradient Descent Search) which is built upon algorithm MG, and its search performance analysis.

Gis2web and its inter-application communications issues

Gis2web is a system intended to bridge the gap between GIS software and datasets and the WWW. The system operates by periodically accessing distributed datasets of GIS information and creating a common repository of WWW accessible information on a gis2web server. In this way gis2web users are provided seamless access to different datasets which may be of quite different forms. This paper discusses the architecture and communications issues of gis2web.

On the Application of BGP Routing Views

In this paper, we present two applications of BGP routing views to the measurement of Web growth and the deployment of IP-based networks or services. A high efficient Importance Sampling approach is introduced to estimate the number of IP addresses of active Web servers, in which the entire IP address space is sampled based on the biasing density obtained from the NEXT-HOP attribute of the BGP data. Additionally the most active AS domains have been identified by exploring the AS-PATH attribute of the BGP routing dumps. This will help discover the approaches to route the IP data in the most efficient way possible.

An Empirical Performance Measurement of Microsoft's Search Engine and its Comparison with Other Major Search Engines

This article reports the results of a project attempting to assess the performance of the new Microsoft search engine (MSE) from various perspectives. Specifically, the study collects statistics, such as the average user response time, average process time for a query reported by MSE itself, the number of pages relevant to a query, and comparisons with its competitors. The project also studies the quality of search results generated by MSE and other search engines using RankPower as the metric. We found MSE per-forms well in speed and diversity of the query results, while weaker in other statistics, compared to some other leading search engines. The contribution of this article is to use different measures to assess and compare the quality of different search engines, especially MSE.