SYSTRA (United Kingdom)

When building machine translation systems, one often needs to make the best out of heterogeneous sets of parallel data in training, and to robustly handle inputs from un-expected domains in testing. This multi-domain scenario has attracted a lot of recent work, that fall under the general umbrella of transfer learning. In this study, we revisit multi-domain machine translation, with the aim to formulate the motivations for developing such systems and the associated expectations with respect to performance. Our experiments with a large sample of multi-domain systems show that most of these expectations are hardly met and suggest that further work is needed to better analyze the current behaviour of multi-domain systems and to make them fully hold their promises.

Abstract The present study examines adults’ processing of the Turkish Aorist, an irregular pattern the acquisition of which is already shown to pose a huge challenge to children up until the age of 8. We investigate how adults produce the Aorist when presented with non-existent monosyllabic roots and to what extent the outcome reflects the application of a rule or the effect of analogy on stored exemplars. Experimental results, associated with a thorough type/token frequency analysis of the aorist in the BOUN corpus, suggest that a rule-based generalization on the basis of the type count of the -Ar affix is at work for over 85% of productions. Similarity-based generalizations, however, though quite few in number appear to be heavily conditioned by the matching last consonant, the matching first and last consonants, overlapping trigram and quadrigram sequences, but not rhyme in Turkish.

Domain adaptation is an old and vexing problem for machine translation systems.The most common and successful approach to supervised adaptation is to fine-tune a baseline system with in-domain parallel data.Standard fine-tuning however modifies all the network parameters, which makes this approach computationally costly and prone to overfitting.A recent, lightweight approach, instead augments a baseline model with supplementary (small) adapter layers, keeping the rest of the model unchanged.This has the additional merit to leave the baseline model intact and adaptable to multiple domains.In this paper, we conduct a thorough analysis of the adapter model in the context of a multidomain machine translation task.We contrast multiple implementations of this idea using two language pairs.Our main conclusions are that residual adapters provide a fast and cheap method for supervised multi-domain adaptation; our two variants prove as effective as the original adapter model and open perspective to also make adapted models more robust to label domain errors.

Recovery time in the rail industry is the additional time that is included in train timetables over and above the minimum journey time necessary often with the explicit aim of improving punctuality. Recovery time is widely used in railways in a number of countries but prior to this study there has been no investigation of the rail users’ point of view. Perceived recovery time, such as being held outside stations and prolonged stops at stations, might have some premium valuation due to the frustration caused. If perceived recovery time in train timetables does carry a premium, then the benefits of improved punctuality achieved by it will be reduced. This paper is the first to investigate passengers’ views and preferences on the use of recovery time. We summarise the findings of a large study and provide estimates of passengers’ valuations of recovery time, both relative to in-vehicle time and late time, that can be used for economic appraisal purposes. Overall, we find most passengers support the use of recovery time but the context is important. Only 13% of users disapprove of its use as a tool to reduce lateness. The estimated premia vary by demand characteristics and are significant in some contexts, although on average are of a small magnitude. The applicability of the estimates is demonstrated through the appraisal of an actual scheme in the UK. We observe that the introduction of more recovery time along with the subsequent improvement in reliability can lead to significant reductions in generalised journey time, even when recovery time carries a valuation premium. We must however strike a word of caution since we note that there were higher than expected proportions of non-traders in the survey which may have affected the results; future studies into the topic should look to minimise the proportion of non-traders. This study provides valuable and necessary first steps in this challenging topic.

Latency is an explicit and possibly insidious issue associated with distributed interactive simulations (DISs), whether they are legacy simulations or simulations based on the DIS protocol or on the HLA (High Level Architecture) and RTI (Run-Time Infrastructure) standards. The implementation details of individual simulations, whether located entirely in one room or geographically distributed across a continent, will directly impact the latencies between various elements of the simulation. At SYSTRAN Corp., we have developed a Windows NT-based system, called SNAPpy (Simulation Network Analysis Project), that seeks to provide tools to both measure and reduce the latency associated with DISs. SNAPpy includes two major sections: an NIU (network interface unit) that is multi-threaded and built on an SMP (symmetric multiprocessor) and Windows NT platform, and the SNAP latency measurement section. This paper reviews the architecture of SNAPpy, going into specific details about its make-up and performance. Finally, SNAPpy's potential as a translator of legacy simulations into the HLA/RTI domain is described, and we introduce the SNAP-Lite Latency Measurement System, a new SYSTRAN product.