Domain-specific Tasks Research Articles

Graph NLU enabled question answering system

With a huge amount of information being stored as structured data, there is an increasing need for retrieving exact answers to questions from tables. Answering natural language questions on structured data usually involves semantic parsing of query to a machine understandable format which is then used to retrieve information from the database. Training semantic parsers for domain specific tasks is a tedious job and does not guarantee accurate results. In this paper, we used conversational analytics tool to create the user interface and to get the required entities and intents from the query thus avoiding the traditional semantic parsing approach. We then make use of Knowledge Graph for querying in structured data domain. Knowledge graphs can be easily leveraged for question answering systems, to use them as the database. We extract appropriate answers for different types of queries which have been illustrated in the Results section.

A span-based joint model for extracting entities and relations of bacteria biotopes.

Information about bacteria biotopes (BB) is important for fundamental research and applications in microbiology. BB task at BioNLP-OST 2019 focuses on the extraction of locations and phenotypes of microorganisms from PubMed abstracts and full-text excerpts. The subtask BB-rel+ner aims to recognize relevant entities and extract interrelationships about BBs. The corresponding corpus owns some distinctive features (e.g. nested entities) which are challenging to deal with. Therefore, previous methods achieved low performance on entity and relation extraction and limited the mutual effect between named entity recognition and relation extraction. There is still much room for improvement. We propose a span-based model to extract entities and relations jointly from biomedical text regarding the BBs. For alleviating the problem of annotated data deficiency in domain-specific task, we employ a BERT (Bidirectional Encoder Representations from Transformers) model pre-trained on the domain-specific corpus to encode sentences. Our model considers all spans in a sentence as potential entity mentions and computes relation scores between the most confident entity spans based on representations of spans and contexts between spans. Experiments on the BB-rel+ner 2019 corpus demonstrate that our model achieves significantly better performance than the state-of-the-art method, with a reduction of 21.6% slot error rate (SER) for extracting relations. Our model is also effective in recognizing nested entities. Furthermore, the model can be applied to the CHEMPROT corpus for joint extraction of chemical-protein entities and relations, achieving state-of-the-art performance. Our source code is available at https://github.com/zmmzGitHub/SpanMB_BERT. Supplementary data are available at Bioinformatics online.

Cognitive Impairment After COVID-19-A Review on Objective Test Data.

Objective: The aim was to conduct a review on the literature on objective cognitive impairment in patients after COVID-19.Methods: We performed a literature review and searched Ovid Medline in February 2021 based on a PECO scheme.Results: Twelve articles met all inclusion criteria. Total patient sample was <1,000. All studies on global cognitive function found impairment, ranging from 15 to 80% of the sampled patients. Seven studies on attention and executive functions reported impairment, with varying results depending on sub-domain and different tests. Three out of four studies reported memory difficulties, with two studies reporting short-term memory deficits. Although results indicate possible language impairment, only one study used domain-specific language tasks. Two out of four studies on visuospatial function did not report any impairment.Conclusion: Patients with recent SARS-CoV-2 infection appear to experience global cognitive impairment, impairment in memory, attention and executive function, and in particular verbal fluency. Based on the current results, we recommend clinicians to evaluate the need for cognitive assessment of patients with a recent COVID-19 infection, regardless of the severity of the disease, treatment methods and length of ICU stay. We need studies with larger sample and control group.

Domain-general and domain-specific influences on emerging numerical cognition: Contrasting uni-and bidirectional prediction models

Domain-general skills such as executive functions (EFs), and domain-specific skills such as non-symbolic number sense and symbolic understanding are often pitted against each other as predictors of emerging maths. Here we aimed to investigate early childhood relations between these foundational skills with a balanced, longitudinal design. One hundred and seventy 3- and 4-year-old-children were tested at two time points, 5 months apart, on four domain-general executive and five domain-specific numeracy tasks. A latent EF factor was a strong predictor of symbolic maths and of their growth. In addition, stronger symbolic maths at Time 1 was correlated with later stronger EF, but symbolic maths did not predict EF growth. Our findings provide novel insights into dynamic interplay between general and specific cognitive skills contributing to preschool maths.

MenuNER: Domain-Adapted BERT Based NER Approach for a Domain with Limited Dataset and Its Application to Food Menu Domain

Entity-based information extraction is one of the main applications of Natural Language Processing (NLP). Recently, deep transfer-learning utilizing contextualized word embedding from pre-trained language models has shown remarkable results for many NLP tasks, including Named-entity recognition (NER). BERT (Bidirectional Encoder Representations from Transformers) is gaining prominent attention among various contextualized word embedding models as a state-of-the-art pre-trained language model. It is quite expensive to train a BERT model from scratch for a new application domain since it needs a huge dataset and enormous computing time. In this paper, we focus on menu entity extraction from online user reviews for the restaurant and propose a simple but effective approach for NER task on a new domain where a large dataset is rarely available or difficult to prepare, such as food menu domain, based on domain adaptation technique for word embedding and fine-tuning the popular NER task network model ‘Bi-LSTM+CRF’ with extended feature vectors. The proposed NER approach (named as ‘MenuNER’) consists of two step-processes: (1) Domain adaptation for target domain; further pre-training of the off-the-shelf BERT language model (BERT-base) in semi-supervised fashion on a domain-specific dataset, and (2) Supervised fine-tuning the popular Bi-LSTM+CRF network for downstream task with extended feature vectors obtained by concatenating word embedding from the domain-adapted pre-trained BERT model from the first step, character embedding and POS tag feature information. Experimental results on handcrafted food menu corpus from customers’ review dataset show that our proposed approach for domain-specific NER task, that is: food menu named-entity recognition, performs significantly better than the one based on the baseline off-the-shelf BERT-base model. The proposed approach achieves 92.5% F1 score on the YELP dataset for the MenuNER task.

Automatic Generation of the Draft Procuratorial Suggestions Based on an Extractive Summarization Method: BERTSLCA

The automatic generation of the draft procuratorial suggestions is to extract the description of illegal facts, administrative omission, description of laws and regulations, and other information from the case documents. Previously, the existing deep learning methods mainly focus on context-free word embeddings when addressing legal domain-specific extractive summarization tasks, which cannot get a better semantic understanding of the text and in turn leads to an adverse summarization performance. To this end, we propose a novel deep contextualized embeddings-based method BERTSLCA to conduct the extractive summarization task. The model is mainly based on the variant of BERT called BERTSUM. Firstly, the input document is fed into BERTSUM to get sentence-level embeddings. Then, we design an extracting architecture to catch the long dependency between sentences utilizing the Bi-Long Short-Term Memory (Bi-LSTM) unit, and at the end of the architecture, three cascaded convolution kernels with different sizes are designed to extract the relationships between adjacent sentences. Last, we introduce an attention mechanism to strengthen the ability to distinguish the importance of different sentences. To the best of our knowledge, this is the first work to use the pretrained language model for extractive summarization tasks in the field of Chinese judicial litigation. Experimental results on public interest litigation data and CAIL 2020 dataset all demonstrate that the proposed method achieves competitive performance.

A combined recall and rank framework with online negative sampling for Chinese procedure terminology normalization.

Medical terminology normalization aims to map the clinical mention to terminologies coming from a knowledge base, which plays an important role in analyzing electronic health record and many downstream tasks. In this article, we focus on Chinese procedure terminology normalization. The expressions of terminology are various and one medical mention may be linked to multiple terminologies. Existing studies based on learning to rank does not fully consider the quality of negative samples during model training and the importance of keywords in this domain-specific task. We propose a combined recall and rank framework to solve these problems. A pair-wise Bert model with deep metric learning is used to recall candidates. Previous methods either train Bert in a point-wise way or based on a multi-class classification problem, which may lead serious efficiency problems or not be effective enough. During model training, we design a novel online negative sampling algorithm to activate the pair-wise method. To deal with multi-implication scenarios, we train the task of implication number prediction together with the recall task in a multi-task learning setting, since these two tasks are highly complementary. In rank step, we propose a keywords attentive mechanism to focus on domain-specific information such as procedure sites and procedure types. Finally, a fusion block merges the results of the recall and the rank model. Detailed experimental analysis shows our proposed framework has a remarkable improvement on both performance and efficiency. The source code will be available at https://github.com/sxthunder/CMTN upon publication.

Knowledge-aware Named Entity Recognition with Alleviating Heterogeneity

Named Entity Recognition (NER) is a fundamental and important research topic for many downstream NLP tasks, aiming at detecting and classifying named entities (NEs) mentioned in unstructured text into pre-defined categories. Learning from labeled data only is far from enough when it comes to domain-specific or temporally-evolving entities (medical terminologies or restaurant names). Luckily, open-source Knowledge Bases (KBs) (Wikidata and Freebase) contain NEs that are manually labeled with predefined types in different domains, which is potentially beneficial to identify entity boundaries and recognize entity types more accurately. However, the type system of a domain-specific NER task is typically independent of that of current KBs and thus exhibits heterogeneity issue inevitably, which makes matching between the original NER and KB types (Person in NER potentially matches President in KBs) less likely, or introduces unintended noises without considering domain-specific knowledge (Band in NER should be mapped to Out_of_Entity_Types in the restaurant-related task). To better incorporate and denoise the abundant knowledge in KBs, we propose a new KB-aware NER framework (KaNa), which utilizes type-heterogeneous knowledge to improve NER. Specifically, for an entity mention along with a set of candidate entities that are linked from KBs, KaNa first uses a type projection mechanism that maps the mention type and entity types into a shared space to homogenize the heterogeneous entity types. Then, based on projected types, a noise detector filters out certain less-confident candidate entities in an unsupervised manner. Finally, the filtered mention-entity pairs are injected into a NER model as a graph to predict answers. The experimental results demonstrate KaNa's state-of-the-art performance on five public benchmark datasets from different domains.

Dirty Pixels: Towards End-to-end Image Processing and Perception

Real-world, imaging systems acquire measurements that are degraded by noise, optical aberrations, and other imperfections that make image processing for human viewing and higher-level perception tasks challenging. Conventional cameras address this problem by compartmentalizing imaging from high-level task processing. As such, conventional imaging involves processing the RAW sensor measurements in a sequential pipeline of steps, such as demosaicking, denoising, deblurring, tone-mapping, and compression. This pipeline is optimized to obtain a visually pleasing image. High-level processing, however, involves steps such as feature extraction, classification, tracking, and fusion. While this silo-ed design approach allows for efficient development, it also dictates compartmentalized performance metrics without knowledge of the higher-level task of the camera system. For example, today’s demosaicking and denoising algorithms are designed using perceptual image quality metrics but not with domain-specific tasks such as object detection in mind. We propose an end-to-end differentiable architecture that jointly performs demosaicking, denoising, deblurring, tone-mapping, and classification (see Figure 1). The architecture does not require any intermediate losses based on perceived image quality and learns processing pipelines whose outputs differ from those of existing ISPs optimized for perceptual quality, preserving fine detail at the cost of increased noise and artifacts. We show that state-of-the-art ISPs discard information that is essential in corner cases, such as extremely low-light conditions, where conventional imaging and perception stacks fail. We demonstrate on captured and simulated data that our model substantially improves perception in low light and other challenging conditions, which is imperative for real-world applications such as autonomous driving, robotics, and surveillance. Finally, we found that the proposed model also achieves state-of-the-art accuracy when optimized for image reconstruction in low-light conditions, validating the architecture itself as a potentially useful drop-in network for reconstruction and analysis tasks beyond the applications demonstrated in this work. Our proposed models, datasets, and calibration data are available at https://github.com/princeton-computational-imaging/DirtyPixels .

Automatic prediction of COVID- 19 from chest images using modified ResNet50.

Recently coronavirus 2019 (COVID-2019), discovered in Wuhan city of China in December 2019 announced as world pandemic by the World Health Organization (WHO). It has catastrophic impacts on daily lives, public health, and the global economy. The detection of coronavirus (COVID− 19) is now a critical task for medical specialists. Laboratory methods for detecting the virus such as Polymerase Chain Reaction, antigens, and antibodies have pros and cons represented in time required to obtain results, accuracy, cost and suitability of the test to phase of infection. The need for accurate, fast, and cheap auxiliary diagnostic tools has become a necessity as there are no accurate automated toolkits available. Other medical investigations such as chest X−ray and Computerized Tomography scans are imaging techniques that play an important role in the diagnosis of COVID− 19 virus. Application of advanced artificial intelligence techniques for processing radiological imaging can be helpful for the accurate detection of this virus. However, Due to the small dataset available for COVID− 19, transfer learning from pre−trained convolution neural networks, CNNs can be used as a promising solution for diagnosis of coronavirus. Transfer learning becomes an effective mechanism by transferring knowledge from generic object recognition tasks to domain-specific tasks. Hence, the main contribution of this paper is to exploit the pre−trained deep learning CNN architectures as a cornerstone to enhance and build up an automated tool for detection and diagnosis of COVID− 19 in chest X−Ray and Computerized Tomography images. The main idea is to make use of their convolutional neural network structure and its learned weights on large datasets such as ImageNet. Moreover, a modification to ResNet50 is proposed to classify the patients as COVID infected or not. This modification includes adding three new layers, named, ′Conv′, ′Batch_Normaliz′ and ′Activation_Relu′ layers. These layers are injected in the ResNet50 architecture for accurate discrimination and robust feature extraction. Extensive experiments are carried out to assess the performance of the proposed model on COVID− 19 chest X−Ray and Computerized Tomography scan images. Experimental results approve that the proposed modification, injected layers, increases the diagnosis accuracy to 97.7% for Computerized Tomography dataset and 97.1% for X−Ray dataset which is superior compared to other approaches.

Heterogeneity in Risk-Taking During the COVID-19 Pandemic: Evidence From the UK Lockdown.

In two pre-registered online studies during the COVID-19 pandemic and the early 2020 lockdown (one of which with a UK representative sample) we elicit risk-tolerance for 1,254 UK residents using four of the most widely applied risk-taking tasks in behavioral economics and psychology. Specifically, participants completed the incentive-compatible Balloon Analog Risk Task (BART) and the Binswanger-Eckel-Grossman (BEG) multiple lotteries task, as well as the Domain-Specific Risk-Taking Task (DOSPERT) and the self-reported questions for risk-taking used in the German Socio-economic Panel (SOEP) study. In addition, participants in the UK representative sample answered a range of questions about COVID-19-related risky behaviors selected from the UCL COVID-19 Social Survey and the ICL-YouGov survey on COVID-19 behaviors. Consistently with pre-COVID-19 times, we find that risk tolerance during the UK lockdown (i) was higher in men than in women and (ii) decreased with age. Undocumented in pre-COVID-19 times, we find some evidence for healthier participants displaying significantly higher risk-tolerance for self-reported risk measures. We find no systematic nor robust patterns of association between the COVID-19 risky behaviors and the four risk-taking tasks in our study. Moreover, we find no evidence in support of the so-called “risk compensation” hypothesis. If anything, it appears that participants who took greater risk in real-life COVID-19-relevant risky behaviors (e.g., isolating or taking precautions) also exhibited higher risk-tolerance in our experimental and self-reported risk-taking measures.

The Relationship between Theory of Mind and Intelligence: A Formative g Approach.

Theory of Mind (ToM) is the ability understand that other people’s mental states may be different from one’s own. Psychometric models have shown that individual differences in ToM can largely be attributed to general intelligence (g) (Coyle et al. 2018). Most psychometric models specify g as a reflective latent variable, which is interpreted as a general ability that plays a causal role in a broad range of cognitive tasks, including ToM tasks. However, an alternative approach is to specify g as a formative latent variable, that is, an overall index of cognitive ability that does not represent a psychological attribute (Kovacs and Conway 2016). Here we consider a formative g approach to the relationship between ToM and intelligence. First, we conducted an SEM with reflective g to test the hypothesis that ToM is largely accounted for by a general ability. Next, we conducted a model with formative g to determine whether the relationship between ToM and intelligence is influenced by domain-specific tasks. Finally, we conducted a redundancy analysis to examine the contribution of each g variable. Results suggest that the relationship between ToM and intelligence in this study was influenced by language-based tasks, rather than solely a general ability.

How does information consistency influence prospective teachers’ decisions about task difficulty assignments? A within-subject experiment to explain data-based decision-making in heterogeneous classes

In order to provide adaptive instruction in heterogeneous classes, teachers have to consider performance-related information for their decisions on which task difficulty level fits best for a particular student. Focusing on prospective teachers’ data-based decision-making, we aim to investigate their use of proximal and distal indicators of students’ ability and readiness to deal successfully with domain-specific tasks that are available for 32 student cases in a digital test environment. To address several hypotheses on the participants’ information-processing behaviors and decisions, we conducted a within-subject experiment in which we systematically varied the informational consistency of the presented student cases. We used a mixed-methods approach to measure observable information-processing behavior and decision-making. We assessed the effects of consistent/inconsistent cases on the amount, sequence, and perceived relevance of selected information, the type of processed information, the time needed for information processing, and the subjective confidence when making a decision via linear dynamic panel-data modeling.

Creative-digital diagnostics and reflexive development of school students in the course of studying Natural Sciences

The authors of the article raises an issue that is relevant from the perspective of modern education connected with difficulties of using the results of psychological diagnostics of students’ creative thinking carried out with the help of traditional testing techniques within the framework of the actual learning process. To address this issue, it is recommended to use the solutions suggested by the “forming” and “reflexive” approaches to this problem and conduct a synthesis of these two approaches using modern digital technologies. This way, it will provide us with an opportunity to turn developing (forming) procedures into diagnostic and turn diagnostic means into developing. Such a conceptual and technological symbiosis allowed us to create a computer networking system “CREO DATUM” enabling its users to carry out creative-digital diagnostics and psycho-pedagogical support of reflexive development of students’ creative potential at the same time based on the material of one of the natural sciences taught within the school curriculum, namely chemistry. This system includes a database of domain-specific creative tasks by solving which students demonstrate their thinking abilities; digitalization of this process in the online mode allows teachers to monitor and assess the dynamics of such parameters of their creative potential as divergence and criticality. This tool enables the teacher and student to predict the area of future development promptly and select further creative tasks on the subject via “CREO DATUM” taking into consideration the strengths and weaknesses of previous creative efforts. Experimental testing of the “CREO DATUM” system allowed us to categorize the students who took part in the research. Four groups have been identified: children with high creative potential, ones with low creative potential, divergents, and analysts. Methodological recommendations have been developed for each category of students based on the material of a natural science subject aimed at psychological and pedagogical support of their development.

Clarifying the role of higher-level cortices in resolving perceptual ambiguity using ultra high field fMRI

The brain is organized into distinct, flexible networks. Within these networks, cognitive variables such as attention can modulate sensory representations in accordance with moment-to-moment behavioral requirements. These modulations can be studied by varying task demands; however, the tasks employed are often incongruent with the postulated functions of a sensory system, limiting the characterization of the system in relation to natural behaviors. Here we combine domain-specific task manipulations and ultra-high field fMRI to study the nature of top-down modulations. We exploited faces, a visual category underpinned by a complex cortical network, and instructed participants to perform either a stimulus-relevant/domain-specific or a stimulus-irrelevant task in the scanner. We found that 1. perceptual ambiguity (i.e. difficulty of achieving a stable percept) is encoded in top-down modulations from higher-level cortices; 2. the right inferior-temporal lobe is active under challenging conditions and uniquely encodes trial-by-trial variability in face perception.

Expert Programmers Have Fine-Tuned Cortical Representations of Source Code.

Expertise enables humans to achieve outstanding performance on domain-specific tasks, and programming is no exception. Many studies have shown that expert programmers exhibit remarkable differences from novices in behavioral performance, knowledge structure, and selective attention. However, the underlying differences in the brain of programmers are still unclear. We here address this issue by associating the cortical representation of source code with individual programming expertise using a data-driven decoding approach. This approach enabled us to identify seven brain regions, widely distributed in the frontal, parietal, and temporal cortices, that have a tight relationship with programming expertise. In these brain regions, functional categories of source code could be decoded from brain activity and the decoding accuracies were significantly correlated with individual behavioral performances on a source-code categorization task. Our results suggest that programming expertise is built on fine-tuned cortical representations specialized for the domain of programming.

Task dependent deep LDA pruning of neural networks

With deep learning’s success, a limited number of popular deep nets have been widely adopted for various vision tasks. However, this usually results in unnecessarily high complexities and possibly many features of low task utility. In this paper, we address this problem by introducing a task-dependent deep pruning framework based on Fisher’s Linear Discriminant Analysis (LDA). The approach can be applied to convolutional, fully-connected, and module-based deep network structures, in all cases leveraging the high decorrelation of neuron motifs found in the pre-decision space and cross-layer deconv dependency. Moreover, we examine our approach’s potential in network architecture search for specific tasks and analyze the influence of our pruning on model robustness to noises and adversarial attacks. Experimental results on datasets of generic objects (ImageNet, CIFAR100) as well as domain specific tasks (Adience, and LFWA) illustrate our framework’s superior performance over state-of-the-art pruning approaches and fixed compact nets (e.g. SqueezeNet, MobileNet). The proposed method successfully maintains comparable accuracies even after discarding most parameters (98%–99% for VGG16, up to 82% for the already compact InceptionNet) and with significant FLOP reductions (83% for VGG16, up to 64% for InceptionNet). Through pruning, we can also derive smaller, but more accurate and more robust models suitable for the task.

Cognitive Evaluation of Machine Learning Agents

Advances in applying statistical Machine Learning (ML) led to several claims of human-level or near-human performance in tasks such as image classification & speech recognition. Such claims are unscientific primarily for two reasons, (1) They incorrectly enforce the notion that task-specific performance can be treated as manifestation of General Intelligence and (2) They are not verifiable as currently there is no set benchmark for measuring human-like cognition in a machine learning agent. Moreover, ML agent’s performance is influenced by knowledge ingested in it by its human designers. Therefore, agent’s performance may not necessarily reflect its true cognition. In this paper, we propose a framework that draws parallels from human cognition to measure machine’s cognition. Human cognitive learning is quite well studied in developmental psychology with frameworks and metrics in place to measure actual learning. To either believe or refute the claims of human-level performance of machine learning agent, we need scientific methodology to measure its cognition. Our framework formalizes incremental implementation of human-like cognitive processes in ML agents with an implicit goal to measure it. The framework offers guiding principles for measuring, (1) Task-specific machine cognition and (2) General machine cognition that spans across tasks. The framework also provides guidelines for building domain-specific task taxonomies to cognitively profile tasks. We demonstrate application of the framework with a case study where two ML agents that perform Vision and NLP tasks are cognitively evaluated.

The face image super-resolution algorithm based on combined representation learning

Face super-resolution reconstruction is the process of predicting high-resolution face images from one or more observed low-resolution face images, which is a typical pathological problem. As a domain-specific super-resolution task, we can use facial priori knowledge to improve the effect of super-resolution. We propose a method of face image super-resolution reconstruction based on combined representation learning method, using deep residual networks and deep neural networks as generators and discriminators, respectively. First, the model uses residual learning and symmetrical cross-layer connection to extract multilevel features. Local residual mapping improves the expressive capability of the network to enhance performance, solves gradient dissipation in network training, and reduces the number of convolution cores in the model through feature reuse. The feature expression of the face image at the high-dimensional visual level is obtained. The visual feature is sent to the decoder through the cross-layer connection structure. The deconvolution layer is used to restore the spatial dimension gradually and repair the details and texture features of the face. Finally, combine the attention block and the residual block reconstruction in the deep residual network to super-resolution face images that are highly similar to high-resolution images and difficult to be discriminated by the discriminator. On this basis, combined representation learning is conducted to obtain numerous realistic results of visual perception. The experimental results on the face datasets can show that the Peak Signal-to-Noise Ratio of the proposed method is improved.

ProSy: API-Based Synthesis with Probabilistic Model

Program synthesis is an exciting topic that desires to generate programs satisfying user intent automatically. But in most cases, only small programs for simple or domain-specific tasks can be synthesized. The major obstacle of synthesis lies in the huge search space. A common practice in addressing this problem is using a domain-specific language, while many approaches still wish to synthesize programs in general programming languages. With the rapid growth of reusable libraries, component-based synthesis provides a promising way, such as synthesizing Java programs which are only composed of APIs (application programming interfaces). However, the efficiency of searching for proper solutions for complex tasks is still a challenge. Given an unfamiliar programming task, programmers would search for API usage knowledge from various coding resources to reduce the search space. Considering this, we propose a novel approach named ProSy to synthesize API-based programs in Java. The key novelty is to retrieve related knowledge from Javadoc and Stack Overflow and then construct a probabilistic reachability graph. It assigns higher probabilities to APIs that are more likely to be used in implementing the given task. In the synthesis process, the program sketch with a higher probability will be considered first; thus, the number of explored reachable paths would be decreased. Some extension and optimization strategies are further studied in the paper. We implement our approach and conduct several experiments on it. We compare ProSy with SyPet and other state-of-the-art API-based synthesis approaches. The experimental results show that ProSy reduces the synthesis time of SyPet by up to 80%.