Reordering Model Research Articles

Statistical machine translation for Indic languages

Abstract Statistical Machine Translation (SMT) systems use various probabilistic and statistical Natural Language Processing (NLP) methods to automatically translate from one language to another language while retaining the originality of the context. This paper aims to discuss the development of bilingual SMT models for translating English into fifteen low-resource Indic languages (ILs) and vice versa. The process to build the SMT model is described and explained using a workflow diagram. Samanantar and OPUS corpus are utilized for training, and Flores200 corpus is used for fine-tuning and testing purposes. The paper also highlights various preprocessing methods used to deal with corpus noise. The Moses open-source SMT toolkit is being investigated for the system’s development. The impact of distance-based reordering and Morpho-syntactic Descriptor Bidirectional Finite-State Encoder (msd-bidirectional-fe) reordering on ILs is compared in the paper. This paper provides a comparison of SMT models with Neural Machine Translation (NMT) for ILs. All the experiments assess the translation quality using standard metrics such as BiLingual Evaluation Understudy, Rank-based Intuitive Bilingual Evaluation Score, Translation Edit Rate, and Metric for Evaluation of Translation with Explicit Ordering. From the result, it is observed that msd-bidirectional-fe reordering performs better than the distance-based reordering model for ILs. It is also noticed that even though the IL-English and English-IL systems are trained using the same corpus, the former performs better for all the evaluation metrics. The comparison between SMT and NMT shows that across various languages, SMT performs better in some cases, while NMT outperforms in others.

The thermal history of Permian carbonate strata reconstructed with clumped isotopes and U–Pb dating: Eastern Sichuan Basin, SW China

The lack of effective paleogeothermometers restricts the study of the thermal history of carbonate strata. Carbonate clumped isotope thermometry and in-situ U–Pb dating are appropriate for reconstructing the thermal history of carbonate strata. Herein, this study analyzed clumped isotope compositions and U–Pb ages of calcite fabrics, including micrite matrix and veins, from Permian carbonate strata in the eastern Sichuan Basin. U–Pb dating of the micrite matrix was performed to rule out the influence of recrystallization and determine the diagenetic age of the samples. Clumped isotope temperatures of the four micrite matrix samples ranged from 102 °C to 137 °C. The maximum paleotemperature derived from a solid-state reordering model was between 210 °C and 220 °C. From all four samples, three calcite veins yielded crystallization temperatures corresponding to 160 ± 10 °C, 60 ± 10 °C, and 29 ± 4 °C, as determined by clumped isotope and solid-state reordering model analyses. In-situ U–Pb dating provided corresponding crystallization ages of 159 ± 17 Ma, 256.6 ± 5.5 Ma, and 17.86 ± 0.8 Ma. The crystallization ages and temperatures of the three calcite veins align well with the thermal history of the Permian strata, indicating a high degree of accuracy. This study introduces a new strategy for reconstructing the thermal history of carbonate strata. The thermal history can be quantified using the constraints of the crystallization ages and temperatures from different stages of calcite veins or cements. This approach is suitable for determining the maturation progress of the source rock and conducting oil and gas exploration in petroliferous basins.

Carbonate clumped isotopes and blocking temperatures of marbles from the Backbone Range, Taiwan

Clumped isotopic compositions (abundance of 13C16O18O in CO2, expressed by Δ47) in massive calcite marble (CM) and dolomite marble (DM) were reported to refine our understanding of greenschist-facies metamorphism during the still on-going arc-continent collision from the northern, central, and southern localities along the northern-central Tailuko belt of the Backbone Range of Taiwan. Black calcite marble (BCM), white calcite marble (WCM), and DM lenses with cataclastic texture were analyzed. We observed that the carbonate clumped isotope reordering is independent of the grain size of minerals and is less likely affected by lattice defects and active stress. The averaged Δ47 values for CMs and DMs are 0.308 ± 0.004 ‰ (n = 26) and 0.235 ± 0.003 ‰ (n = 13), respectively, giving blocking temperatures of 211 ± 6 °C and 332 ± 8 °C. The blocking temperatures infer mean cooling rates of 614 °C/m.y. and 8.8 × 104 °C/m.y for calcites and dolomites, respectively, estimated based on the available models and clumped reordering mechanisms. We noted that the variations in cooling rates widely differ with different reordering models, especially with different temperature calibrations. Some cooling rate estimates are in agreement with those based on zircon/apatite thermochronological studies. Further refinements on Δ47 - temperature calibration, reordering mechanism and estimation of thermodynamic constants, in addition to improvement in analytical precision, are needed for practical applications of carbonate clumped thermometry as a tectonic speedometer for metamorphic terranes.

Lexical Based Reordering Models for English to Telugu Machine Translation

Lexical Based Reordering Models for English to Telugu Machine Translation

Joint application of fluid inclusion and clumped isotope (∆47) thermometry unravels the complexity of thermal and fluid‐flow history reconstruction in sedimentary basins—Upper Triassic Chaunoy Formation reservoirs (Paris Basin)

AbstractWe investigated calcites and dolomites precipitated during burial diagenesis of the Upper Triassic (Norian) continental siliciclastics from sub‐surface reservoirs of the northern Paris Basin (Chaunoy Formation) that experienced a thermal maximum >100°C during Late Cretaceous times. Relative carbonate precipitation timing was established via petrographic analyses. The diagenetic carbonates were further investigated by fluid inclusion and clumped isotope (Δ47) thermometry. The two thermometric datasets were interpreted by evaluating the possible occurrence of inclusion thermal reequilibration and Δ47 solid‐state reordering, based on the known basin thermal history and the three existing Δ47 reordering models. By considering the fluid inclusion and Δ47 datasets obtained and the various Δ47 reordering models, different carbonate precipitation scenarios, in terms of timing and parent fluid composition (δ18Ofluid), were inferred. These results underline that in samples having experienced thermal maximum >100°C, accuracy and interpretation of fluid inclusion and Δ47 thermometry data (especially on calcite) may be biased by thermal reequilibration and solid‐state reordering. The results converge towards the need of jointly applying fluid inclusion and Δ47 thermometry on the same carbonate phases to evaluate all the possible precipitation scenarios. The most likely carbonate precipitation scenarios, based on Δ47 thermometry data, point at the precipitation of two calcite phases during Early to Late Jurassic times and of one dolomite phase during the Late Cretaceous. The parent fluids possibly were original formation waters of the Chaunoy Fm. that mixed with brines migrating from the East, where time equivalent evaporitic deposits occur. The proposed precipitation model for calcites and dolomites, involving different pulses of brine migration, and the dominance of calcite phases were not recorded by previous studies on the Upper Triassic units. These latter results may be of interest to evaluate the reservoir potential of the Chaunoy Fm. in this underexplored portion of the Paris Basin.

Efficient SNN multi-cores MAC array acceleration on SpiNNaker 2.

The potential low-energy feature of the spiking neural network (SNN) engages the attention of the AI community. Only CPU-involved SNN processing inevitably results in an inherently long temporal span in the cases of large models and massive datasets. This study introduces the MAC array, a parallel architecture on each processing element (PE) of SpiNNaker 2, into the computational process of SNN inference. Based on the work of single-core optimization algorithms, we investigate the parallel acceleration algorithms for collaborating with multi-core MAC arrays. The proposed Echelon Reorder model information densification algorithm, along with the adapted multi-core two-stage splitting and authorization deployment strategies, achieves efficient spatio-temporal load balancing and optimization performance. We evaluate the performance by benchmarking a wide range of constructed SNN models to research on the influence degree of different factors. We also benchmark with two actual SNN models (the gesture recognition model of the real-world application and balanced random cortex-like network from neuroscience) on the neuromorphic multi-core hardware SpiNNaker 2. The echelon optimization algorithm with mixed processors realizes 74.28% and 85.78% memory footprint of the original MAC calculation on these two models, respectively. The execution time of echelon algorithms using only MAC or mixed processors accounts for ≤ 24.56% of the serial ARM baseline. Accelerating SNN inference with algorithms in this study is essentially the general sparse matrix-matrix multiplication (SpGEMM) problem. This article explicitly expands the application field of the SpGEMM issue to SNN, developing novel SpGEMM optimization algorithms fitting the SNN feature and MAC array.

Thermally-induced clumped isotope resetting in belemnite and optical calcites: Towards material-specific kinetics

The application of carbonate clumped isotope (Δ47) thermometry in deep-time is often limited by modification of the original temperature signal by thermal resetting. New modeling approaches to estimate the initial isotopic composition of partially reset calcites and maximal burial temperatures, however, open promising avenues in temperature reconstruction. Such approaches strongly depend on laboratory-derived kinetic parameters of calcite materials, which may differ in their microstructure, water content and distribution, and minor and trace element composition, and thus may have different resetting kinetics. The rostra of belemnites, an extinct group of mollusks with a wide temporal and spatial occurrence in the Mesozoic, have been extensively used for deep-time paleoclimate reconstructions using oxygen isotope geochemistry. Belemnites are also important targets for clumped isotope-based temperature reconstructions, but often are found to have reset Δ47 compositions. Here, we present results from heating experiments on belemnite rostral calcite and optical calcite and provide belemnite-specific kinetic parameters for clumped isotope resetting. We show that belemnite calcite is altered faster and at lower temperatures than optical calcite and all other calcites reported in previous studies. We suggest that fast initial resetting results from oxygen isotope exchange of belemnite calcite with internal skeletal water present as fluid inclusions or organic-derived water, a process completed within 2–4 min at the experimental temperatures used here. Extrapolation to geological timescales using different solid-state bond reordering models shows that belemnite calcite resetting starts at lower burial temperatures than brachiopod, spar, and optical calcites. This susceptibility to thermal resetting results in a measurable (+3 °C) increase of the apparent Δ47 temperature even under shallow to moderate burial conditions (i.e., 40–50 °C for 106–107 years timescales). Following the overprint to higher apparent Δ47 temperatures during burial, the belemnite Δ47 may further re-equilibrate during exhumation resulting in a decrease of apparent Δ47 temperatures. Such “retrograde resetting” is similar to what is observed for carbonatites and marbles during cooling, and may cause underestimation of the thermal resetting a sample experienced during its geological history. Overall, our results demonstrate the importance of material-specific kinetic parameters and we urge caution when interpreting Δ47-derived temperatures of biogenic carbonates from deep-time archives.

Clumped isotope analysis of zoned calcite cement, Carboniferous, Isle of Man

AbstractSequential analyses of δ13C, δ18O and Δ47 values of calcite and dolomite deposited in millimetre‐sized cavities are reported from the Ronaldsway Member packstones, Isle of Man. The Ronaldsway brachiopods have δ13C values of ca +2.3‰ and δ18O values of ca −7.2‰; carbon is like predicted Carboniferous values, while oxygen values are more negative. The brachiopods show preserved microstructure but have marginal alteration and a streaky cathodoluminescence pattern. Crinoid ossicles have δ13C values of ca +2.3‰ and one with a δ18O value of ca −3.1‰, compatible with Carboniferous marine precipitates; three samples have δ18O values of ca −6.5‰ and are 18O‐depleted. Calcite stages 1 and 2 have δ13C values ca +3.2‰ and δ18O values ca −2.5‰, compatible with Carboniferous sea water. Stage 1 and 2 have non‐luminescent to orange CL zones. Stage 1 and early stage 2 contain red luminescent dolomite micro crystals generated during Mg calcite stabilisation. The Δ47 values for stage 1 and 2 cements indicate temperatures of 86 and 105°C that occurred after the stabilisation of Mg calcite. Stage 3–8 zoned cements preserve their original growth surfaces and their δ13C and δ18O values suggest precipitation during burial and exhumation. The Δ47 values of the brachiopods and crinoids indicate temperatures between 85 and 140°C indicating they were either recrystallised at high temperatures or affected by solid state reordering. To evaluate these alternatives two quantitative models, water–rock reaction and reordering models are performed. The allochems and cements are progressively altered by porewater towards the fluid‐buffered behaviour. The quantitative evaluation of calcite and dolomite solid‐state reordering suggests the elevated clumped isotopic temperatures are produced by interaction with hydrothermal fluids. This study improves understanding by applying previously untried techniques; further Δ47 data and quantifying elemental variations would help further interpretation but the poorly documented post‐depositional history is a drawback.

Clumped isotopes of paired dolomite and calcite constraining alteration histories of ancient carbonate successions

Clumped isotope thermometry is an effective tool which allows to understand the post-depositional alteration, including recrystallization and solid-state reordering, of carbonate rocks. However, as both processes are controlled by elevated temperatures, it is frequently a challenge to distinguish between the relative importance of these two processes in deeply buried carbonate successions (depth > 5000 m or formation temperature > 150°C). In this paper, we have established a clumped isotopic framework of paired dolomite and calcite in order to decipher the relative influence of solid-state reordering versus recrystallization and consequently offer an improved understanding of the mechanisms involved. The material studied was derived from two deeply buried successions (Jurassic-Ediacaran) drilled in the central Sichuan Basin of South China. The calcite and dolomite minerals from these cores were separated using buffered acetic acid, and based on a two-component mixing model, the carbon, oxygen and clumped isotopic (Δ47) compositions of the two endmembers were determined. Based on an evaluation of solid-state reordering models, we conclude that micritic carbonates (dolomite and calcite) from the Jurassic to the Cambrian have been altered during burial by the geothermal gradient as determined by vitrinite reflectance. The Ediacaran micritic dolomite records a hydrothermal event in the transition between the Ediacaran and Cambrian. In contrast, the clumped isotopic temperature of Ediacaran saddle dolomites reflects subsequent hydrothermal activity during the Silurian and Devonian and shows compositional variations controlled by cooling rates. This study highlights the utility of the theoretical framework based on paired dolomite and calcite records to untangle overprinted signals of clumped isotopes in ancient carbonate rocks and gives insight to better understand the complexity of alteration processes in deeply burial carbonate successions.

Partitioning and Reordering for Spike-Based Distributed-Memory Parallel Gauss--Seidel

Gauss--Seidel (GS) is a widely used iterative method for solving sparse linear systems of equations and also known to be effective as a smoother in algebraic multigrid methods. Parallelization of GS is a challenging task since solving the sparse lower triangular system in GS constitutes a sequential bottleneck at each iteration. We propose a distributed-memory parallel GS (dmpGS) by implementing a parallel sparse triangular solver (stSpike) based on the Spike algorithm. stSpike decouples the global triangular system into smaller systems that can be solved concurrently and requires the solution of a much smaller reduced sparse lower triangular system which constitutes a sequential bottleneck. In order to alleviate this bottleneck and to reduce the communication overhead of dmpGS, we propose a partitioning and reordering model consisting of two phases. The first phase is a novel hypergraph partitioning model whose partitioning objective simultaneously encodes minimizing the reduced system size and the communication volume. The second phase is an in-block row reordering method for decreasing the nonzero count of the reduced system. Extensive experiments on a dataset consisting of 359 sparse linear systems verify the effectiveness of the proposed partitioning and reordering model in terms of reducing the communication and the sequential computational overheads. Parallel experiments on 12 large systems using up to 320 cores demonstrate that the proposed model significantly improves the scalability of dmpGS.

Reassessment of ocean paleotemperatures during the Late Ordovician

Abstract The Late Ordovician is a paradox, with a greenhouse climate shifting to an icehouse climate during a time of presumably high CO2 levels. We used clumped isotope (Δ47) microanalysis of fossil brachiopod shells to determine paleotemperatures of North America (Cincinnati Arch) during the Katian (453–443 Ma). Microanalyses of fossil brachiopods yielded a right-skewed distribution of clumped isotope temperatures [T(Δ47)] ranging from 25 °C to 55 °C with a mode of 32 °C and a mean of 36 °C. To test the impact of Δ47 reordering (reequilibration of 13C-18O bonding) at burial temperatures on T(Δ47), we applied a Monte Carlo simulation to a Δ47 reordering model based on the burial history. Because the reordering simulation indicated partial reordering, we used the simulation results to back-calculate T(Δ47) and correct the reordering effect. Correcting for reordering decreased the mean T(Δ47) by only ~1 °C, to 35 °C, and did not change the mode (32 °C). These temperatures are lower than those of previous studies, but still suggest that the Late Ordovician climate of the Cincinnati Arch was warmer than the modern subtropical seas. However, carbonate facies in these strata argue for upwelling of cool water; thus, typical subtropical surface waters may have been as warm as 35 °C approaching the Hirnantian glaciation. Seawater δ18O values derived from back-calculated T(Δ47) and brachiopod δ18O values average −0.3‰ ± 0.6‰ relative to Vienna standard mean ocean water (VSMOW), consistent with the value expected for subtropical surface waters in a greenhouse Earth, assuming a constant seawater-δ18O model.

Joint Learning With BERT-GCN and Multi-Attention for Event Text Classification and Event Assignment

Government hotline is closely related to people’s lives and plays an important role in solving social problems and maintaining social stability in China. However, the event text of the hotline is inconsistent in length and unclear in elements, so it is a challenge for the operator to manually complete the assignment tasks of hotline event. To address these problems, we propose a joint learning method for event text classification and event assignment for Chinese government hotline. Firstly, graph convolution network (GCN) and BERT are used to process the event text respectively to obtain the corresponding representation vector. Then, the obtained two representation vectors are fused by the dynamic fusion gate to get fusion vector and classified the fusion vector through the text classification. Secondly, we use multi-attention mechanism to process the GCN result vector, BERT result vector and the “sanding” vector to obtain attentive “event-sanding” representation vector and calculate the corresponding department probability distribution. Finally, the historical prior knowledge based reorder model is used to sort the results of the “event-sanding” matching module and output the optimal assignment department of government hotline event. Experimental results show that our method can achieve better performance compared with several baseline approaches. The ablation experiments also demonstrate the validity of each proposed module in our model.

Empirical Analysis of Phrase-Based Statistical Machine Translation System for English to Hindi Language

Hindi is the national language of India. However, most of the Government records, resolutions, news, etc. are documented in English which remote villagers may not understand. This fact motivates to develop an automatic language translation system from English to Hindi. Machine translation is the process of translating a text in one natural language into another natural language using computer system. Grammatical structure of Hindi language is very much complex than English language. The structural difference between English and Hindi language makes it difficult to achieve good quality translation results. In this paper, phrase-based statistical machine translation approach (PBSMT) is used for translation. Translation, reordering and language model are main working components of a PBSMT system. This paper evaluates the impact of various combinations of these PBSMT system parameters on automated English to Hindi language translation quality. Freely available n-gram-based BLEU metric and TER metric are used for evaluating the results.

Thermal Alteration History of the Fenghuoshan Group, Hoh Xil Basin, Northern Tibetan Plateau: Insights From Clumped Isotope Thermometry

AbstractThe carbonate clumped isotope (Δ47) thermometer has been widely used for paleoclimatic and paleo‐elevation reconstructions in sedimentary basins. Its reliable use requires that carbonates are free from later diagenetic alteration and retain their depositional Δ47 compositions, which is not always the case during the burial history of a basin. Therefore, identifying the alteration mechanisms that occurred during burial is essential for the correct interpretation of Δ47 signatures. In this study, we used the Δ47 thermometer to investigate possible carbonate alteration during burial of the Fenghuoshan Group (FHSG) in the Hoh Xil Basin, northern Tibetan Plateau, China. The Δ47 temperatures [T(Δ47)] of 30 carbonate samples through the ∼4.5 km thickness of the FHSG range from 51 to 132°C and exceed the surface temperature, implying the effects of thermal alteration on FHSG carbonates. The T(Δ47) values combined with the water‐rock exchange model, indicating that the alteration occurred under closed burial conditions at low water:rock ratios. The solid‐state reordering model fails to predict the Δ47 values, even with a modified time‐temperature history. In contrast, most measured Δ47 values are in agreement with predictions from recrystallization models. Therefore, the proposed alteration scenario is that FHSG carbonates underwent progressive microscale recrystallization during deep burial and subsequent exhumation, during which process recrystallized carbonate equilibrated with ambient burial temperature and therefore modified Δ47 values. This study highlights the potential of Δ47 thermometry and related models for evaluating the burial history and diagenesis of carbonates in sedimentary basins.

Using Clumped Isotopes to Reconstruct the Maximum Burial Temperature: A Case Study in the Sichuan Basin

For strata that have experienced continual burial in the early stage and uplift in the late stage, the present-day temperature is lower than the maximum burial temperature (MBT), which is a key parameter for studying the hydrocarbon generation history of source rocks in petroliferous basins. In this paper, a new method for reconstructing the MBT is proposed based on the solid-state reordering model of carbonate clumped isotopes (Δ47). The MBT reconstructed using the Δ47 was compared with the MBT constrained using the traditional Easy%Ro model. The clumped isotope temperature (TΔ47) of the Permian micritic limestone from the Xibeixiang outcrop (about 62°C) is much higher than its initial formation temperature (20–25°C), suggesting that the limestone experienced partial solid-state reordering during the late burial process. The MBT of the calcite obtained from the solid-state reordering model is 139–147°C, which is quite similar to the MBT determined using the Easy%Ro model (139.5–147.5°C). TΔ47 of the Permian and Triassic limestone and calcite cements in the Puguang gas field are 150–180°C, while TΔ47 of the micritic dolostone is about 70°C, suggesting that the Δ47 of the limestone and calcite cements experienced complete solid-state reordering and the dolostone only experienced partial solid-state reordering. The MBT of the dolomite determined using the solid-state reordering model is 200–220°C, which is also similar to the MBT determined using the Easy%Ro model (202–227°C). Therefore, the case studies from the Sichuan Basin suggest that Δ47 can be used to reconstruct the MBT of ancient carbonate strata lacking vitrinite and detrital zircon data. However, different types of carbonate samples should be used to reconstruct the MBT for strata that have experienced different temperature histories. Micritic limestone and very finely crystallized dolostone can be used to reconstruct the MBT of strata that have experienced MBTs of <150–200°C and >200–250°C, respectively.

Source-side Reordering to Improve Machine Translation between Languages with Distinct Word Orders

English and Hindi have significantly different word orders. English follows the subject-verb-object (SVO) order, while Hindi primarily follows the subject-object-verb (SOV) order. This difference poses challenges to modeling this pair of languages for translation. In phrase-based translation systems, word reordering is governed by the language model, the phrase table, and reordering models. Reordering in such systems is generally achieved during decoding by transposing words within a defined window. These systems can handle local reorderings, and while some phrase-level reorderings are carried out during the formation of phrases, they are weak in learning long-distance reorderings. To overcome this weakness, researchers have used reordering as a step in pre-processing to render the reordered source sentence closer to the target language in terms of word order. Such approaches focus on using parts-of-speech (POS) tag sequences and reordering the syntax tree by using grammatical rules, or through head finalization. This study shows that mere head finalization is not sufficient for the reordering of sentences in the English-Hindi language pair. It describes various grammatical constructs and presents a comparative evaluation of reorderings with the original and the head-finalized representations. The impact of the reordering on the quality of translation is measured through the BLEU score in phrase-based statistical systems and neural machine translation systems. A significant gain in BLEU score was noted for reorderings in different grammatical constructs.

A Thermal Pulse Induced by a Permian Mantle Plume in the Tarim Basin, Northwest China: Constraints From Clumped Isotope Thermometry and In Situ Calcite U‐Pb Dating

AbstractHigh burial temperature and a lack of material for traditional thermal proxies significantly impede the understanding of the thermal history of carbonate stratigraphic sequences in deep time. Carbonate clumped isotope thermometry interpreted through models of solid‐state isotopic reordering and in situ U‐Pb dating provides a new methodology for reconstructing thermal histories of such successions. Calcite fabrics (micrite cements and veins) from deeply buried (5,000–8,000 m) Ordovician carbonate intervals in the Tarim Basin were analyzed for their carbonate clumped isotope compositions and U‐Pb ages, and the results were used to estimate maximum peak burial temperatures. Reconstructed burial temperatures reveal a previously ignored thermal pulse in the range of 160°C–190°C during the Permian period. Maximum geothermal gradients at the time of peak burial temperature varied from 26.8°C/km to 69.3°C/km, with higher values in the northern Tarim Basin. We show that the spatial patterns of peak burial temperature and geothermal gradients closely correspond to the distribution of the Tarim Large Igneous Province, which emplaced during the Permian following the emplacement of a mantle plume beneath the Tarim Basin. We thus propose that the thermal pulse recorded in Tarim Ordovician limestones resulted from Permian mantle plume activity. The reconstructed thermal history has important implications for our understanding of source‐rock hydrocarbon generating history, hydrocarbon accumulation, and fluid flow history in the Tarim Basin. This study also demonstrates the potential of the integration of Δ47, U‐Pb dating of carbonates and solid‐state reordering models as an integrated thermochronological method for reconstructing thermal histories of deeply buried carbonate intervals.

Model and Simulation of Maximum Entropy Phrase Reordering of English Text in Language Learning Machine

This paper proposes a feature extraction algorithm based on the maximum entropy phrase reordering model in statistical machine translation in language learning machines. The algorithm can extract more accurate phrase reordering information, especially the feature information of reversed phrases, which solves the problem of imbalance of feature data during maximum entropy training in the original algorithm, and improves the accuracy of phrase reordering in translation. In the experiment, they were combined with linguistic features such as parts of speech, words, and syntactic features extracted by using the syntax analyzer, and the maximum entropy classifier was used to predict translation errors, and the experimental verification was performed on the Chinese-English translation data set and compared. The experimental results show that different word posterior probabilities have a significant impact on the classification error rate, and the combination of linguistic features based on the word posterior probability can significantly reduce the classification error rate and improve the translation error prediction performance.

Combination of Δ47 and U-Pb dating in tectonic calcite veins unravel the last pulses related to the Pyrenean Shortening (Spain)

Clumped isotopes thermometry (Δ47) and U-Pb dating of carbonates have been used to elucidate the temperature-time conditions of 10 tectonic calcite veins in the hanging wall of the South Pyrenean Frontal Thrust (Spain), the youngest thrust unit in the south-central Pyrenees. The Δ47 values indicate precipitation temperatures below ∼90°C, in agreement with fluid inclusion analyses. 6 veins have been successfully dated, giving ages between 61 and 14.5 Ma with all but one value below 24 Ma. On the basis of our new results, we propose that several the studied veins record the end of Pyrenean tectonics in the area during the Burdigalian, possibly up to the Serravalian, younger than estimates based on magnetostratigraphic data from continental syn-tectonic deposits. Calculated δ18O values of the precipitating fluid are either very positive, suggesting a local source that has interacted with the host rock, or negative, suggesting the downward incursion of meteoric waters. For undated or pre-burial veins, the Δ47 values derived from temperature history reordering model (THRM) show that the measured Δ47 values may have been significantly altered during burial, preventing their use as a reliable thermal or paleohydrological marker. Although applied to a limited number of samples, these results highlight the great potential of the U-Pb/Δ47 approach to decipher the history of vein or cement formation and fluid flow in complex tectonic zones, provided that (i) the thermal evolution is known and (ii) the extent of clumped isotope reordering is systematically modelled.

Reconstructing the magnitude of Early Toarcian (Jurassic) warming using the reordered clumped isotope compositions of belemnites

The magnitude of temperature changes in the Early Jurassic are not well known. Clumped isotope measurements can potentially be used to provide better constrains, but unfortunately many of the well-studied sedimentary successions that preserve Lower Jurassic fossils experienced burial temperatures above the limits of preservation of Δ47, which for geological timescales is thought to be between 80–120 °C. Samples from these basins are expected to be partially reordered and yield apparent clumped isotope temperatures that are warmer than original values. Here, we explore whether useful paleoclimate information can be recovered from these samples. We test the hypothesis that relative temperature differences are preserved in partially reordered samples when they experience a common burial history. This was done with the use of reordering models and Δ47 measurements of early Jurassic belemnites from the Aubach section of the SW German Basin, a basin that has a relatively well constrained burial history with maximum burial temperatures above 90 °C. We find that even though partial reordering progressively erases the Δ47 difference between samples, the majority (>50%) of the signal is preserved when samples are buried at temperatures as high of 150 °C for up to 200 Ma. Moreover, the models demonstrate that – regardless of burial conditions – partially reordered samples always preserve minimum records of temperature change across climate events. These inferences are supported by the belemnite Δ47 data that show partially reordered compositions and warming/cooling patterns across the Early Jurassic that closely mimic observations from independent proxies. Model observations are used to interpret a 13 ± 4 °C (95% ci) temperature increase that is observed in the belemnite data across the Early Toarcian. The large magnitude of the temperature excursion is explained as a combination of warming and a change in belemnite habitat before and after the Toarcian Ocean Anoxic Event. Our results demonstrate the usefulness of partially reordered samples and further open the use of this proxy in deep time settings.