The effects of fitness functions on genetic programming-based ranking discovery for Web search

Findings on ranking evaluation functions for feature weighting in image retrieval

Background

There are substantial benefits to be gained from ranking optimization in several information retrieval and recommendation systems. However, the analysis of ranking evaluation functions (REFs), which play a major role in many ranking optimization models, needs to be further investigated. An analysis of previous studies that investigated REFs was performed, and evidence was found which indicated that the choice of a proper REF is context sensitive.

Methods

In this study, we analyze a broad set of REFs for feature weighting aimed at increasing the image retrieval effectiveness. The REFs analyzed sums ten and includes the most successful and representative REFs from the literature. The REFs were embedded into a genetic algorithm (GA)-based relevance feedback (RF) model, called WLSP-C ±, aimed at improving image retrieval results through the use of learning weights for image descriptors and image regions.

Results

Analyses of precision-recall curves in five real-world image data sets showed that one non-parameterized REF named F5, not analyzed in previous studies, overcame recommended ones, which require parameter adjustment. We also provided a computational analysis of the GA-based RF model investigated, and it was shown that it is linear in regard to the image data set cardinality.

Conclusions

We conclude that REF F5 should be investigated in other contexts and problem scenarios centered on ranking optimization, as ranking optimization techniques rely heavily on the ranking quality measure.

IR with and without GA

This paper proposes two IR approaches; the first is IR with GA, which is a GA-based IR approach. This approach introduces modified GA operators that allow IR with GA to achieve high performance. The second IR model is IR without GA, which is based on traditional IR approach. Both enhance the precision and recall of the web search by improving the document representation where an enhanced inverted index is developed for this purpose. Moreover, these two models use the same proposed evaluation function for measuring the document relativity to the user query. A number of experiments were conducted to compare the performance of the two suggested approaches with existing techniques. The two suggested approaches were then compared experimentally with another two techniques of classical IR namely Okapi-BM25 fitness function and Bayesian inference network model from documents quality of retrieval perspective. The obtained results demonstrate a good level of enhancement to the recall and precision times. In addition, the documents retrieved by IR with and without GA are more accurate and relevant to the queries than that retrieved by other techniques. Overall, the two suggested approaches provide a promising technique in web search domain delivering a high quality search results in terms of recall and precision.