In this paper, we present a fast and strong neural approach for general purpose text matching applications. We explore what is sufficient to build a fast and well-performed text matching model and propose to keep three key features available for inter-sequence alignment: original point-wise features, previous aligned features, and contextual features while simplifying all the remaining components. We conduct experiments on four well-studied benchmark datasets across tasks of natural language inference, paraphrase identification and answer selection. The performance of our model is on par with the state-of-the-art on all datasets with much fewer parameters and the inference speed is at least 6 times faster compared with similarly performed ones.

No abstract

Fine-grained sentiment analysis is a useful tool for producers to understand consumers’ needs as well as complaints about products and related aspects from online platforms. In this article, we define a novel task named “Multi-Entity Aspect-Based Sentiment Analysis (ME-ABSA)”. It investigates the sentiment towards entities and their related aspects. It makes the well-studied aspect-based sentiment analysis a special case of this type, where the number of entities is limited to one. We contribute a new dataset for this task, with multi-entity Chinese posts in it. We propose to model context, entity, and aspect memory to address the task and incorporate dependency information for further improvement. Experiments show that our methods perform significantly better than baseline methods on datasets for both ME-ABSA task and ABSA task. The in-depth analysis further validates the effectiveness of our methods and shows that our methods are capable of generalizing to new (entity, aspect) combinations with little loss of accuracy. This observation indicates that data annotation in real applications can be largely simplified.

Inspired by recent works in Aspect-Based Sentiment Analysis (ABSA) on product reviews and faced with more complex posts on social media platforms mentioning multiple entities as well as multiple aspects, we define a novel task called Multi-Entity Aspect-Based Sentiment Analysis (ME-ABSA). This task aims at fine-grained sentiment analysis of (entity, aspect) combinations, making the well-studied ABSA task a special case of it. To address the task, we propose an innovative method that models Context memory, Entity memory and Aspect memory, called CEA method. Our experimental results show that our CEA method achieves a significant gain over several baselines, including the state-of-the-art method for the ABSA task, and their enhanced versions, on datasets for ME-ABSA and ABSA tasks. The in-depth analysis illustrates the significant advantage of the CEA method over baseline methods for several hard-to-predict post types. Furthermore, we show that the CEA method is capable of generalizing to new (entity, aspect) combinations with little loss of accuracy. This observation indicates that data annotation in real applications can be largely simplified.

As important biomolecules in Camellia sinensis L., amino acids (AAs) are considered to contribute to the overall green tea sensory quality and undergo dynamic changes during growth. However, limited by analytical capacity, detailed AAs composition in different growth stages remains unclear. To address this question, we analyzed the dynamic changes of 23 AAs during leaf growth in Xinyang Mao Jian (XYMJ) green tea. Using amino acid analyzer, we demonstrated that most AAs are abundant on Pure Brightness Day and Grain Rain Day. After Grain Rain, 23 AAs decreased significantly. Further analysis shows that theanine has a high level on the day before Spring Equinox and Grain Rain, accounting for 44–61% of the total free AAs content in tea leaves. Glu, Pro, and Asp are the second most abundant AAs. Additionally, spinasterol and 22,23-dihydrospinasterol are first purified and identified in ethanol extract of XYMJ by silica gel column chromatography method. This study reveals the relationship between plucking days and the dynamic changes of AAs during the growth stage and proves the rationality of the traditional plucking days of XYMJ green tea.

Single molten track (SMT) is the basic unit of the selective laser melting (SLM) process, of which the morphology features and microstructures have great significance on the molding quality. In this study, the effects of laser energy density (LED) on characteristics of SMT were studied with 316 L stainless steel processing by SLM. The results indicate that an optimized range of LED was between 200 and 500 J/mm3, which was the best for SLM process of 316 L leading to beautiful SMT appearance such as continuous uniform water-ridge-like surface, good contour straightness, smooth edges, little adhesive powder, and high densification. The microstructures of the SMT molten pool are mainly composed of equiaxed cellular and columnar dendrites. As the LED decreased, the cellular became smaller and disappeared, while the length and width of columnar crystals even gradually increased. The size of the cellular varies between 0.43 and 1.70 μm with the LEDs increase. The relationship between SMT characteristics and LED found in this work can be readily used to optimize SMT and control the microstructures in situ, and the laser energy density range that the work has identified will simplify most of the guess work in additive manufacturing.

In this study, we aim to reduce generation latency for Named Entity Recognition (NER) with Large Language Models (LLMs). The main cause of high latency in LLMs is the sequential decoding process, which autoregressively generates all labels and mentions for NER, significantly increase the sequence length. To this end, we introduce Parallel Decoding in LLM for NER (PaDeLLM-NER), a approach that integrates seamlessly into existing generative model frameworks without necessitating additional modules or architectural modifications. PaDeLLM-NER allows for the simultaneous decoding of all mentions, thereby reducing generation latency. Experiments reveal that PaDeLLM-NER significantly increases inference speed that is 1.76 to 10.22 times faster than the autoregressive approach for both English and Chinese. Simultaneously it maintains the quality of predictions as evidenced by the performance that is on par with the state-of-the-art across various datasets.* Code is available at URL masked for anonymous review.* https://huggingface.co/meta-llama/Llama-2-7b * https://huggingface.co/baichuan-inc/ Baichuan2-7B-Base