LKMT:Linguistics Knowledge-Driven Multi-Task Neural Machine Translation for Urdu and English

Thanks to the strong representation capability of pre-trained language models,supervised machine translation models have achieved outstanding performance.However,the performances of these models drop sharply when the scale of the parallel training corpus is limited.Considering the pre-trained language model has a strong ability for monolingual representation,it is the key challenge for machine translation to construct the in-depth relationship between the source and target language by injecting the lexical and syntactic information into pre-trained language models.To alleviate the dependence on the parallel corpus,we propose a Linguistics Knowledge-Driven MultiTask(LKMT)approach to inject part-of-speech and syntactic knowledge into pre-trained models,thus enhancing the machine translation performance.On the one hand,we integrate part-of-speech and dependency labels into the embedding layer and exploit large-scale monolingual corpus to update all parameters of pre-trained language models,thus ensuring the updated language model contains potential lexical and syntactic information.On the other hand,we leverage an extra self-attention layer to explicitly inject linguistic knowledge into the pre-trained language model-enhanced machine translation model.Experiments on the benchmark dataset show that our proposed LKMT approach improves the Urdu-English translation accuracy by 1.97 points and the English-Urdu translation accuracy by 2.42 points,highlighting the effectiveness of our LKMT framework.Detailed ablation experiments confirm the positive impact of part-of-speech and dependency parsing on machine translation.

A PROPORTIONAL 3-TUPLE FUZZY LINGUISTIC REPRESENTATION MODEL FOR SCREENING NEW PRODUCT PROJECTS

It is critically important for companies to screen new product projects before they are launched to the market. So far, many approaches have been developed for tackling the process of screening product innovations. Due to uncertain, vague and incomplete information as well as dynamically complex process regarding to new product development （NPD）, a fuzzy linguistic approach employed linguistic assessments and the fuzzy-set-based computation is reasonable for screening new products. However, such a fuzzy linguistic approach faces with various defects and limitations, such as loss of information, failing in considering the aspects related to human nature on uncertain subjective judgments etc. These defects and limitations lead to a dilemma, i.e., it＇s very difficult to screen new product projects reasonably and precisely. In this paper, we propose a notion of proportional 3-tuple to represent a linguistic assessment and related ignoring information, and a preference-preserving proportional 3-tuple transformation for the unification of linguistic assessments represented by proportional 3-tuples between two different linguistic term sets. On this basis, a proportional 3-tuple fuzzy linguistic representation model for screening new product projects is developed. It is shown that the proposed model is flexible to handle uncertain, vague and incomplete information related to screening new product projects. It not only allows evaluators to express their subjective judgments with different confidence levels, but is also able to deal with incomplete linguistic assessments. Ultimately, the proposed model also improves the precision and reasonability of the screening result.

A Web-Based Integrated Intelligent System for Sensory Fabric Hand Evaluation

This paper presents a web-based integrated system for on-line sensory fabric hand evaluation. The methods of fuzzy techniques, neural networks, classical factorial analysis and other data analysis are used in the system to analyze the objective and subjective data, and to build the relationship between them. Given the objective data of a new fabric sample, the system can provide its sensory hand data and its total hand grade. In meantime, the total hand grade can be obtained directly from the sensory fabric hand data if provided. The sensory evaluation system is developed in Internet environment using Java language and SQL server database management system.

高校英语语言学课程动态评价模式研究

作为一名英语专业学生必须要学的一门课程就是英语语言学。现今英语语言学教学因该课程评价方式过于单一,致使该课程的学习不能满足大纲提出的标准。怎样对英语语言学评价进行合理有效的改革就变得异常关键。合理有效的评价体系是保障教学成功的关键因素,因此,可以从多元化的角度评价,如:评价主体、评价内容、评价方式、评价手段,尝试探讨适合时代需要的动态评价模式。

Evaluating Emergency Response Capacity of Chemical Industrial Park Using an Hybrid Fuzzy AHP Method

This paper presents an expert-based fuzzy analytic hierarchy process( AHP) model for evaluating emergency response capacity of Chemical Industrial Park( ERCCIP) by jointly using an improved fuzzy preference programming( FPP) and 2-tuple fuzzy linguistic approach. An evaluation index system for ERCCIP is proposed. The weight of sub-criteria and criteria of the evaluation index system for ERCCIP are determined using the improved FPP. And the ratings of sub-criteria are assessed in linguistic values according to the experts' subjective opinions. Finally,the aggregated ratings of criteria and the overall ERCCIP are calculated.