arXiv subjects

Xiaodong Shi

Publications and source records attributed to Xiaodong Shi.

50 records · Page 3Linked to original sources

arXiv2023

LEAPT: Learning Adaptive Prefix-to-prefix Translation For Simultaneous Machine Translation

Simultaneous machine translation, which aims at a real-time translation, is useful in many live scenarios but very challenging due to the trade-off between accuracy and latency. To achieve the balance for both, the model needs to wait for appropriate streaming text (READ policy) and then generates its translation (WRITE policy). However, WRITE policies of previous work either are specific to the method itself due to the end-to-end training or suffer from the input mismatch between training and decoding for the non-end-to-end training. Therefore, it is essential to learn a generic and better WRITE policy for simultaneous machine translation. Inspired by strategies utilized by human interpreters and "wait" policies, we propose a novel adaptive prefix-to-prefix training policy called LEAPT, which allows our machine translation model to learn how to translate source sentence prefixes and make use of the future context. Experiments show that our proposed methods greatly outperform competitive baselines and achieve promising results.

cs.CL↗

arXiv2023

Sensitivity of $CP$ Violation of $Λ$ decay in $J/ψ\to Λ\barΛ$ at STCF

The process of $J/ψ\to Λ\barΛ$ is studied using $1.0\times10^{12}$ $J/ψ$ Monte Carlo (MC) events at $\sqrt{s}$=3.097 GeV with a fast simulation software at future Super Tau Charm Facility (STCF). The statistical sensitivity for $CP$ violation is determined to be the order of $\mathcal{O} (10^{-4})$ by measuring the asymmetric parameters of the $Λ$ decay. Furthermore, the decay of $J/ψ\to Λ\barΛ$ also serves as a benchmark process to optimize the detector responses using the interface provided by the fast simulation software.

hep-ph↗

arXiv2022

Towards Better Document-level Relation Extraction via Iterative Inference

Document-level relation extraction (RE) aims to extract the relations between entities from the input document that usually containing many difficultly-predicted entity pairs whose relations can only be predicted through relational inference. Existing methods usually directly predict the relations of all entity pairs of input document in a one-pass manner, ignoring the fact that predictions of some entity pairs heavily depend on the predicted results of other pairs. To deal with this issue, in this paper, we propose a novel document-level RE model with iterative inference. Our model is mainly composed of two modules: 1) a base module expected to provide preliminary relation predictions on entity pairs; 2) an inference module introduced to refine these preliminary predictions by iteratively dealing with difficultly-predicted entity pairs depending on other pairs in an easy-to-hard manner. Unlike previous methods which only consider feature information of entity pairs, our inference module is equipped with two Extended Cross Attention units, allowing it to exploit both feature information and previous predictions of entity pairs during relational inference. Furthermore, we adopt a two-stage strategy to train our model. At the first stage, we only train our base module. During the second stage, we train the whole model, where contrastive learning is introduced to enhance the training of inference module. Experimental results on three commonly-used datasets show that our model consistently outperforms other competitive baselines.

cs.CL↗

arXiv2022

Detecting and analyzing missing citations to published scientific entities

Proper citation is of great importance in academic writing for it enables knowledge accumulation and maintains academic integrity. However, citing properly is not an easy task. For published scientific entities, the ever-growing academic publications and over-familiarity of terms easily lead to missing citations. To deal with this situation, we design a special method Citation Recommendation for Published Scientific Entity (CRPSE) based on the cooccurrences between published scientific entities and in-text citations in the same sentences from previous researchers. Experimental outcomes show the effectiveness of our method in recommending the source papers for published scientific entities. We further conduct a statistical analysis on missing citations among papers published in prestigious computer science conferences in 2020. In the 12,278 papers collected, 475 published scientific entities of computer science and mathematics are found to have missing citations. Many entities mentioned without citations are found to be well-accepted research results. On a median basis, the papers proposing these published scientific entities with missing citations were published 8 years ago, which can be considered the time frame for a published scientific entity to develop into a well-accepted concept. For published scientific entities, we appeal for accurate and full citation of their source papers as required by academic standards.

cs.DL↗

arXiv2022

Automatic Analysis of Available Source Code of Top Artificial Intelligence Conference Papers

Source code is essential for researchers to reproduce the methods and replicate the results of artificial intelligence (AI) papers. Some organizations and researchers manually collect AI papers with available source code to contribute to the AI community. However, manual collection is a labor-intensive and time-consuming task. To address this issue, we propose a method to automatically identify papers with available source code and extract their source code repository URLs. With this method, we find that 20.5% of regular papers of 10 top AI conferences published from 2010 to 2019 are identified as papers with available source code and that 8.1% of these source code repositories are no longer accessible. We also create the XMU NLP Lab README Dataset, the largest dataset of labeled README files for source code document research. Through this dataset, we have discovered that quite a few README files have no installation instructions or usage tutorials provided. Further, a large-scale comprehensive statistical analysis is made for a general picture of the source code of AI conference papers. The proposed solution can also go beyond AI conference papers to analyze other scientific papers from both journals and conferences to shed light on more domains.

cs.SE↗

arXiv2022

Feasibility Study of $D_s^+ \to τ^+ ν_τ$ Decay and Test of Lepton Flavor Universality with Leptonic $D_s^+$ Decays at STCF

We report a sensitive study of $D_s^+ \to τ^+ ν_τ$ decay via $τ^+ \to e^+ ν_e \barν_τ$ with an integrated luminosity of 1 ab$^{-1}$ at the center-of-mass energy of $4.009$ GeV at a future Super Tau Charm Facility (STCF). Under the help of the fast simulation software package, the statistical sensitivity for the absolute branching fraction of $D_s^+ \to τ^+ ν_τ$ is determined to be $2\times10^{-4}$. Combining with our previous prospect of $D_s^+ \to μ^+ ν_μ$, the ratio of the branching fractions for $D_s^+ \to τ^+ ν_τ$ over $D_s^+ \to μ^+ ν_μ$ can achieve a relative statistical precision of 0.5%, which will provide the most stringent test of the $τ$-$μ$ lepton flavor universality in heavy quark decays. Taking the decay constant $f_{D_s^+}$ from lattice QCD calculations or the CKM matrix element $|V_{cs}|$ from the CKMfitter group as an input, the relative statistical uncertainties for $|V_{cs}|$ and $f_{D_s^+}$ are estimated to be 0.3% and 0.2%, respectively.

hep-ex↗

arXiv2021

Prospects of CKM elements $|V_{cs}|$ and decay constant $f_{D_{s}^+}$ in $D_s^+\toμ^+ν_μ$ decay at STCF

We report a feasibility study of pure leptonic decay $D_s^+\toμ^+ν_μ$ by using a fast simulation software package at STCF. With an expected luminosity of $1~\mathrm{ab}^{-1}$ collected at STCF at a center-of-mass energy of 4.009 GeV, the statistical sensitivity of the branching fraction is determined to be 0.3\%. Combining this result with the $c\rightarrow s$ quark mixing matrix element $|V_{cs}|$ determined from the current global Standard Model fit, the statistical sensitivity of $D_s^+$ decay constant, $f_{D_s^+}$, is estimated to be 0.2\%. Alternatively, combining the current results of $f_{D_s^+}$ calculated by lattice QCD, the statistical sensitivity of $|V_{cs}|$ is determined to be 0.2\%, which helps probe possible new physics beyond. The unprecedented precision to be achieved at STCF will provide a precise calibration of QCD and rigorous test of Standard Model.

hep-ex↗

arXiv2021

Feasibility study of $CP$ Violation in $τ^{-}\to K_{S}π^{-}ν_τ$ decays at Super Tau Charm Facility

We report a feasibility study of $CP$ violation of $τ^{-}\rightarrow K_{S}π^{-} ν_τ$ decay at a Super Tau Charm Facility~(STCF).With an expected luminosity of 1~ab$^{-1}$ collected by STCF per year at a center-of-mass energy of 4.26 GeV, the statistical sensitivity for the $CP$ violation is determined to be of order $9.7\times10^{-4}$ by measuring the decay-rate difference between $τ^{+}\rightarrow K_{S}π^{+}\barν_τ$ and $τ^{-}\rightarrow K_{S}π^{-} ν_τ$. The analysis is performed using a reliable fast simulation software package, which can describe the detector responses properly and vary the responses flexibly for further optimization. Moreover, the energy-dependent efficiencies for reconstructing $τ^{-}\rightarrow K_{S}π^{-} ν_τ$ are presented and the expected $CP$ sensitivity is proportional to $1/\sqrt{\mathcal{L}}$ in the energy region from 4.0 to 5.0 GeV. The sensitivity of $CP$ violation is of order $3.1\times10^{-4}$ with 10~ab$^{-1}$ integrated luminosity, which is equivalent to ten years data taking in this energy region at STCF.

hep-ex↗

arXiv2019

Asynchronous "Events" are Better For Motion Estimation

Event-based camera is a bio-inspired vision sensor that records intensity changes (called event) asynchronously in each pixel. As an instance of event-based camera, Dynamic and Active-pixel Vision Sensor (DAVIS) combines a standard camera and an event-based camera. However, traditional models could not deal with the event stream asynchronously. To analyze the event stream asynchronously, most existing approaches accumulate events within a certain time interval and treat the accumulated events as a synchronous frame, which wastes the intensity change information and weakens the advantages of DAVIS. Therefore, in this paper, we present the first neural asynchronous approach to process event stream for event-based camera. Our method asynchronously extracts dynamic information from events by leveraging previous motion and critical features of gray-scale frames. To our best knowledge, this is the first neural asynchronous method to analyze event stream through a novel deep neural network. Extensive experiments demonstrate that our proposed model achieves remarkable improvements against the state-of-the-art baselines.

cs.CV↗

arXiv2019

A complete data processing workflow for CryoET and subtomogram averaging

Electron cryotomography (CryoET) is currently the only method capable of visualizing cells in 3D at nanometer resolutions. While modern instruments produce massive amounts of tomography data containing extremely rich structural information, the data processing is very labor intensive and results are often limited by the skills of the personnel rather than the data. We present an integrated workflow that covers the entire tomography data processing pipeline, from automated tilt series alignment to subnanometer resolution subtomogram averaging. This workflow greatly reduces human effort and increases throughput, and is capable of determining protein structures at state-of-the-art resolutions for both purified macromolecules and cells.

q-bio.QM↗

arXiv2018

Multi-Perspective Neural Architecture for Recommendation System

Currently, there starts a research trend to leverage neural architecture for recommendation systems. Though several deep recommender models are proposed, most methods are too simple to characterize users' complex preference. In this paper, for a fine-grain analysis, users' ratings are explained from multiple perspectives, based on which, we propose our neural architecture. Specifically, our model employs several sequential stages to encode the user and item into hidden representations. In one stage, the user and item are represented from multiple perspectives and in each perspective, the representations of user and item put attentions to each other. Last, we metric the output representations of final stage to approach the users' rating. Extensive experiments demonstrate that our method achieves substantial improvements against baselines.

cs.IR↗

arXiv2018

Hungarian Layer: Logics Empowered Neural Architecture

Neural architecture is a purely numeric framework, which fits the data as a continuous function. However, lacking of logic flow (e.g. \textit{if, for, while}), traditional algorithms (e.g. \textit{Hungarian algorithm, A$^*$ searching, decision tress algorithm}) could not be embedded into this paradigm, which limits the theories and applications. In this paper, we reform the calculus graph as a dynamic process, which is guided by logic flow. Within our novel methodology, traditional algorithms could empower numerical neural network. Specifically, regarding the subject of sentence matching, we reformulate this issue as the form of task-assignment, which is solved by Hungarian algorithm. First, our model applies BiLSTM to parse the sentences. Then Hungarian layer aligns the matching positions. Last, we transform the matching results for soft-max regression by another BiLSTM. Extensive experiments show that our model outperforms other state-of-the-art baselines substantially.

cs.CL↗

arXiv2017

Deep Semantic Role Labeling with Self-Attention

Semantic Role Labeling (SRL) is believed to be a crucial step towards natural language understanding and has been widely studied. Recent years, end-to-end SRL with recurrent neural networks (RNN) has gained increasing attention. However, it remains a major challenge for RNNs to handle structural information and long range dependencies. In this paper, we present a simple and effective architecture for SRL which aims to address these problems. Our model is based on self-attention which can directly capture the relationships between two tokens regardless of their distance. Our single model achieves F$_1=83.4$ on the CoNLL-2005 shared task dataset and F$_1=82.7$ on the CoNLL-2012 shared task dataset, which outperforms the previous state-of-the-art results by $1.8$ and $1.0$ F$_1$ score respectively. Besides, our model is computationally efficient, and the parsing speed is 50K tokens per second on a single Titan X GPU.

cs.CL↗

arXiv2016

Lattice-Based Recurrent Neural Network Encoders for Neural Machine Translation

Neural machine translation (NMT) heavily relies on word-level modelling to learn semantic representations of input sentences. However, for languages without natural word delimiters (e.g., Chinese) where input sentences have to be tokenized first, conventional NMT is confronted with two issues: 1) it is difficult to find an optimal tokenization granularity for source sentence modelling, and 2) errors in 1-best tokenizations may propagate to the encoder of NMT. To handle these issues, we propose word-lattice based Recurrent Neural Network (RNN) encoders for NMT, which generalize the standard RNN to word lattice topology. The proposed encoders take as input a word lattice that compactly encodes multiple tokenizations, and learn to generate new hidden states from arbitrarily many inputs and hidden states in preceding time steps. As such, the word-lattice based encoders not only alleviate the negative impact of tokenization errors but also are more expressive and flexible to embed input sentences. Experiment results on Chinese-English translation demonstrate the superiorities of the proposed encoders over the conventional encoder.

cs.CL↗