The Factors Affecting the Performance of Data Fusion Algorithms

Abstract

The enormous amount of data which is distributed on the World Wide Web can be very useful if the users became able to get these data in an easy and appropriate method, search engines help the users to find what they need from this enormous amount of data. Meta-search is the application of data fusion to document retrieval, Metasearch engine takes as an input the N ranked lists output by each of N search engines in response to a given query, As output, it computes a single ranked list, which is hopefully an improvement over any input list as measured by standard information retrieval performance metrics such as the mean average precision (MAP). Our goal in this paper is to answer the following question, what are the factors affecting the performance of Data fusion algorithms? The reason behind introducing those factors is the absence of a single source in the literature able to present all those factors in an organized and complete manner. This work is needed to integrate all data fusion performance research findings. This paper contributes to the data fusion literature by two things, firstly; it will deliver all factors affecting the performance of data fusion algorithms in an organized and complete manner. Secondly; it will deliver recommendations which are related to how and when to deal with the factors that affect the performance.

Meta-search based approach for Arabic information retrieval

“…Indeed, many data fusion techniques were proposed for other languages; each reporting good results under specific circumstances (Wu, 2012). That is why, some research works tried to find the appropriate fusion model for each situation using statistical analysis (Croft, 2000; Nassar and Kanaan, 2009; Ng and Kantor, 2000). Although many previous data fusion works tried to define the factors affecting fusion performance, each work gave different interpretations.…”

“…Croft (2000) supposed that compatible outputs and independent subsystems are sufficient conditions to get high quality data fusion. Nassar and Kanaan (2009) summarized some factors that affect the performance of data fusion results including new factors such as the average performance of the results. That is why we propose to compare NLP tools to improve fusion in terms of subsystems diversity, independence and quality.…”

Reliability of Measured Data for pH Sensor Arrays with Fault Diagnosis and Data Fusion Based on LabVIEW

“…The faults are diagnosed through three levels which are data fusion level, feature level and decision level respectively. Nassar and Kanaan [ 8 ] surveyed and discussed the state-of-the-art studies related to the factors affecting the performance of data fusion algorithms, and have integrated data fusion performance research findings. Xue [ 9 ] presented a fault diagnosis system based on multi-sensor data fusion algorithm, which is composed of a local data fusion level and whole data fusion level.…”