Báo cáo khoa học: "Query Weighting for Ranking Model Adaptation" pptx

... document in- formation, a probabilistic weighting for ⟨x i , x j ⟩ was modeled by w q ∗ w ij . Through the transformation, instance weighting schemes for classification can be applied to ranking model ... importance weighting for ranking model adapta- tion could be done better at query level rather than document level. Figure 1 demonstrates the difference between in- stance...

Ngày tải lên: 30/03/2014, 21:20

... instances can greatly improve the performance for all tasks. And in almost all cases, weighting the target instances more than the source instances per- formed better than weighting them equally. We also ... approximations in Sec- tion 2.3. Before we show the formal framework, we first introduce some weighting parameters and ex- plain the intuitions behind these parameters. First, for...

Ngày tải lên: 31/03/2014, 01:20

... terms of domain model predicates; its hierarchy is used for enforcing selectional restrictions and for IRule inheritance; and some limited inferencing is done based on the model. After semantic ... interface for Information Retrieval. Proceedings of the 6th Annual International ACM SIGIR Conference, ACM Special Interest Group on Information Retrieval and American Society for...

Ngày tải lên: 08/03/2014, 18:20

... map- ping. Therefore, we obtained new results for these methods in our setup. For the Transfer baseline, for each target language we trained the model on all other languages in our dataset. For the Mixture baseline, ... decreased performance. For this reason, we chose to train the model on the four languages. sider the performance of the supervised version of our model which c...

Ngày tải lên: 16/03/2014, 19:20

... Techniques So far, we discussed mixture modelling for trans- lation models, which is only a subset of domain adaptation techniques in SMT. Mixture-modelling for language models is well established (Foster ... following algorithm. For each value pair (x, y) for which we compute p(x|y), we re- place λ i with 0 for all models i with p(y) = 0, then renormalize the weight vector λ to 1....

Ngày tải lên: 17/03/2014, 22:20

... system’s model. For example, the translation process from one language to an- other language may follow some hidden tree trans- formation process, in a recursive fashion. Many features of the model ... collapsing spurious ambiguity for other tasks as well. Such tasks include data- oriented parsing (DOP), applications of Hidden Markov Models (HMMs) and mixture models, and other models w...

Ngày tải lên: 23/03/2014, 16:21

... be expected from such a simple model, particularly in light of the fact that there is some choice in the word order for most sentences (among adjuncts for exam- ple) and that this model does not take the ... diffi- cult one. One way to break the problem down is to model the DLA as a set of weights for each type of dependency relation. Under this model the word order is determined...

Ngày tải lên: 23/03/2014, 18:20

... appointment scheduling domain is there- fore ideal for a case study aimed at extracting linguistic indicators for definiteness. 3.1 Overview Explicit marking for definiteness takes place on several ... target language. For this paper, we propose a rule hierarchy for this purpose, that can be used as a pre- processing tool to context checking. All noun phrases marked for de...

Ngày tải lên: 23/03/2014, 19:20

... presented a joint approach for template filling that models mention detection, unification, and field extraction in a flexible, feature-rich model. This ap- proach allows for joint modeling of interdependen- cies ... relation, an integrated model to weight the dif- ferent extraction options, including for example the likely lengths for events, or the fact that start times should come b...

Ngày tải lên: 30/03/2014, 17:20

... implying that ranking plays an important role in our approach. Perfect ranking. It would be informative to see whether our ranking model is performing at its up- per limit, because further performance ... apply a perfect ranking model, which uses an oracle to choose the best can- didate partition for each test text. Results in row 7 of Table 3 indicate that our ranking model p...

Ngày tải lên: 20/02/2014, 15:20