Academic
Publications
Relevance weighting for query independent evidence

Relevance weighting for query independent evidence,10.1145/1076034.1076106,Nick Craswell,Stephen E. Robertson,Hugo Zaragoza,Michael J. Taylor

Relevance weighting for query independent evidence   (Citations: 66)
BibTex | RIS | RefWorks Download
A query independent feature, relating perhaps to document content, linkage or usage, can be transformed into a static, per-document relevance weight for use in ranking. The challenge is to find a good function to transform feature values into relevance scores. This paper presents FLOE, a simple density analysis method for modelling the shape of the transformation required, based on training data and without assuming independence between feature and baseline. For a new query independent feature, it addresses the questions: is it required for ranking, what sort of transformation is appropriate and, after adding it, how successful was the chosen transformation? Based on this we apply sigmoid transformations to PageRank, indegree, URL Length and ClickDistance, tested in combination with a BM25 baseline.
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
    • ... cause similar clickrate difference of another two restaurants that are further away from user’s location, e.g., 10 and 11 kilometers instead of 1 and 2 kilometers away? It is critical to understand the underlying behaviors/heuristics of a ranking signal, which would guide us to design more effective ranking features; this has been demonstrated extensively in the development of retrieval functions for general Web search, e.g., [30, 33, 11, ...,It has been previously demonstrated that understanding the behaviors/heuristics of a ranking signal is critical in the development of retrieval functions for Web search [30, 33, 14, 11, 25]...

    Yuanhua Lvet al. An exploration of ranking heuristics in mobile local search

    • ...Incorporating these priors into the scoring function has been demonstrated to improve retrieval on large web collections [11, 16, 28]...

    Michael Benderskyet al. Quality-biased ranking of web documents

    • ...[3]). We experimented with several combination methods, such as using the log of the degrees or prior probabilities trained on the relevance data, instead of the degrees themselves...

    Marijn Koolenet al. Are Semantically Related Links More Effective for Retrieval?

    • ...The approach in Equation 4 has been shown to perform reasonable well in IR literature [15]...
    • ...We have used the linear weighted-log combination approach proposed in [15]...

    Abhay Harpaleet al. CiteData: a new multi-faceted dataset for evaluating personalized sear...

    • ...We consider linear combination of features (transformed with a function as explained in [1])...
    • ...x+� as suggested in [1], where x is the feature to transform and θ is a parameter...

    Gianluca Demartiniet al. Entity summarization of news articles

Sort by: