Analyzing and boosting the performance of explicit result diversification methods for web search

Download

index.pdf

Date

2016

Author

Akçay, Mehmet

Metadata

Show full item record

Item Usage Stats

147
views

84
downloads

A recent study on the topic of additivity addresses the task of search result diversification and concludes that while weaker baselines are usually significantly improved by the evaluated diversification methods, for stronger baselines, no significant improvement can be observed. Due to the importance of the issue in shaping future research evaluation strategies in search results diversification, in this thesis, we first aim to reproduce the findings reported in the previous study, and then investigate its possible limitations. Our extensive experiments first reveal that under the same experimental setting with that previous study, we can reach similar results. Next, we hypothesize that for stronger baselines, tuning the parameters of some methods should be done in a more fine-grained manner. With trade-off parameters that are specifically determined for each baseline run, we show that the percentage of significant improvements even over the strong baselines can be doubled. As a further issue, we discuss the possible impact of using the same strong baseline retrieval function for the diversity computations of the methods. Finally, the effect of another parameter in search result diversification i.e. candidate set size is analyzed and we show that using adaptive candidate set size on a query basis instead of a fixed value across all queries, performances of result diversification methods on strong baselines can be further improved. In conclusion, in the case of a strong baseline, it is more crucial to tune the parameters of the diversification methods to be evaluated; but once this is done, additivity is achievable.

Subject Keywords

Search engines., Web search engines., Search engines

URI

http://etd.lib.metu.edu.tr/upload/12620673/index.pdf
https://hdl.handle.net/11511/26209

Collections

Graduate School of Natural and Applied Sciences, Thesis

Suggestions

OpenMETU
Core

Comparison of non-deterministic search techniques in the optimum design of real size steel frames Hasançebi, Oğuzhan; Doğan, E.; Erdal, F.; Saka, M.P. (Elsevier BV, 2010-9) There is a noticeable increase in the emergence of non-deterministic search techniques that simulate natural phenomena into a numerical optimization technique in recent years. These techniques are used for developing structural optimization algorithms that are particularly effective for obtaining solutions to discrete programming problems. In this study amongst these techniques genetic algorithms, simulated annealing, evolution strategies, particle swarm optimizer, tabu search, ant colony optimization and h...
On the Additivity and Weak Baselines for Search Result Diversification Research Akcay, Mehmet; Altıngövde, İsmail Sengör; Macdonald, Craig; Ounis, Iadh (2017-10-04) A recent study on the topic of additivity addresses the task of search result diversi cation and concludes that while weaker baselines are almost always signi cantly improved by the evaluated diversi cation methods, for stronger baselines, just the opposite happens, i.e., no signi cant improvement can be observed. Due to the importance of the issue in shaping future research directions and evaluation strategies in search results diversi cation, in this work, we rst aim to reproduce the ndings reported in ...
Modeling and implementation of local volatility surfaces in Bayesian framework Animoku, Abdulwahab; Uğur, Ömür; Yolcu-Okur, Yeliz (2018-06-01) In this study, we focus on the reconstruction of volatility surfaces via a Bayesian framework. Apart from classical methods, such as, parametric and non-parametric models, we study the Bayesian analysis of the (stochastically) parametrized volatility structure in Dupire local volatility model. We systematically develop and implement novel mathematical tools for handling the classical methods of constructing local volatility surfaces. The most critical limitation of the classical methods is obtaining negativ...
Analysis of an inventory system under backorder correlated deterministic demand and geometric supply process Argon, NT; Gullu, R; Erkip, N (2001-05-06) In this article we propose a single item, periodic review model that investigates the effects of changes in the demand process that occur after stockout realizations. We investigate a system where the demands in successive periods are deterministic but affected by the backorder realizations. In order to capture the effects of changes in the demand process we use a geometric type supply availability. We analytically derive the necessary components for obtaining profit related performance measures and provide...
Linear Analysis of Zero-Lift Pitching Moment Coefficient and Gravity Effects on Missile Autopilots Ovec, Naz T.; Erer, Koray S.; Kutay, Ali Türker (2015-03-14) The most of the studies on disturbance rejection are covered with robust control methods, where this paper suggests a narrowed but a more simplified linear model based technique to assess the robustness of the controllers. In this paper presented, the problems originated from the disturbances are aimed to be obtained as far in advance as possible. The method used in this work reduces the iterative fine tuning time course between the non-linear simulation and the linear model; hence it enhances the authority...

Citation Formats

M. Akçay, “Analyzing and boosting the performance of explicit result diversification methods for web search,” M.S. - Master of Science, Middle East Technical University, 2016.