Translation model size is growing at a pace that outstrips improvements in computing power, and this hinders research on many interesting models. We show how an algorithmic scaling technique can be used to easily handle very large models. Using this technique, we explore several large model variants and show an improvement 1.4 BLEU on the NIST 2006 ChineseEnglish task. This opens the door for w...

This paper addresses the problem of multiple pattern matching for motifs encoded by Position Weight Matrices. We first present an algorithm that uses a multi-index table to preprocess the set of motifs, allowing a dramatically decrease of computation time. We then show how to take benefit from simlar motifs to prevent useless computations.

The retrieval of images from a large database of images is an important and emerging area of research. Here, a technique to retrieve images based on appearance that works eeectively across large changes of scale is proposed. The database is initially ltered with derivatives of a Gaussian at several scales. A user deened template is then created from an image of an object similar to those being ...

Context matching algorithms automatically discover semantic relations between two autonomously developed conceptual representations of two overlapping domains. Although significant work has been carried on the theoretical ground, the empirical evaluation of matching algorithms is still an open area for research. In this paper we introduce CTXMATCH, a matching algorithm which applies to symbolic...

During the process of citation matching links from bibliography entries to referenced publications are created. Such links are indicators of topical similarity between linked texts, are used in assessing the impact of the referenced document and improve navigation in the user interfaces of digital libraries. In this paper we present a citation matching method and show how to scale it up to hand...

We present a logic-based framework for automated skill matching, able to return a ranked referral list and the related ranking explanation. Thanks to a Knowledge Compilation approach, a knowledge base in Description Logics is translated into a relational database, without loss of information. Skill matching inference services are then efficiently executed via SQL queries. Experimental results f...

Recently, we are witnessing an explosive growth of data in the business and scientific area. In fact, there are many databases and information sources available through the web covering different domains: semantic Web, deep Web, e-business, biology, digital libraries, etc. In such domains, the data generated are heterogeneous and voluminous e.g schemas with several thousand elements are common ...