no code implementations • 29 Jan 2019 • Oosterhuis Harrie, Culpepper J. Shane, de Rijke Maarten
Second, we evaluate the potential gains that can be achieved in terms of memory requirements.
no code implementations • 29 Jan 2019 • Oosterhuis Harrie, de Rijke Maarten
Our findings show that the theoretical bounds of DBGD do not apply to any common ranking model and, furthermore, that the performance of DBGD is substantially worse than PDGD in both ideal and worst-case circumstances.
no code implementations • 16 Jul 2018 • Chen Yifan, de Rijke Maarten
Learning feature representations, on the other hand, ensures a sufficient number of inputs to train a deep network.
no code implementations • 27 Nov 2017 • Xie Xiaohui, Liu Yiqun, de Rijke Maarten, He Jiyin, Zhang Min, Ma Shaoping
(2)How does image search behavior change with user intent?
1 code implementation • 26 Nov 2017 • Oosterhuis Harrie, de Rijke Maarten
Conversely, simpler models can be optimized on fewer interactions and thus provide a better user experience, but they will converge towards suboptimal rankings.
1 code implementation • 26 Nov 2017 • Oosterhuis Harrie, de Rijke Maarten
We show empirically that, compared to previous multileaved comparison methods, PPM is more sensitive to user preferences and scalable with the number of rankers being compared.
no code implementations • 15 Aug 2017 • Li Ziming, Kiseleva Julia, de Rijke Maarten, Grotov Artem
It is natural to represent the behavior of users who are engaging with interactive systems such as a search engine or a recommender system, as a sequence of actions where each next action depends on the current situation and the user reward of taking a particular action.
no code implementations • 14 Jun 2017 • Kiseleva Julia, de Rijke Maarten
Mobile devices differ radically from classic command-based and point-and-click user interfaces, now allowing for gesture-based interaction using fine-grained touch and swipe signals.
no code implementations • 16 Jan 2017 • Azarbonyad Hosein, Dehghani Mostafa, Kenter Tom, Marx Maarten, Kamps Jaap, de Rijke Maarten
We propose a hierarchical re-estimation approach for topic models to combat generality and impurity; the proposed approach operates at three levels: words, topics, and documents.
no code implementations • 14 Sep 2016 • Reinanda Ridho, Meij Edgar, de Rijke Maarten
Results of applying the model to unseen entities are promising, indicating that the model is able to learn the general characteristics of a vital document.
no code implementations • 26 Jan 2015 • Chuklin Aleksandr, de Rijke Maarten
Currently, the quality of a search engine is often determined using so-called topical relevance, i. e., the match between the user intent (expressed as a query) and the content of the document.