لا يزال مخصصات المجال في التحليل النحوي تحديا كبيرا.نحن نعلم مسألة عدم توازن البيانات بين النطاق داخل المجال والخروج من النطاق يستخدم عادة للمشكلة.نحدد تكيف النطاق كمشكلة تعليمية متعددة المهام (MTL)، والتي تتيح لنا تدريب اثنين من المحللين، واحدة لكل منها الرئيسية.تظهر نتائجنا أن نهج MTL مفيد لنقش Treebank الأصغر.بالنسبة لأكبر Treebank، نحتاج إلى استخدام وزن الخسارة من أجل تجنب انخفاض في الأداء المهمة الفردية.من أجل تحديد درجة توهت البيانات، فإن اختلال البيانات بين مجطتين واختلافات المجال تؤثر على النتائج، ونحن نقوم أيضا بتجربة اثنين من Treebanks غير المتوازن داخل المجال وإظهار أن وزن الخسارة يحسن أيضا الأداء في إعداد المجال.نظرا لارتداء الخسارة في MTL، يمكننا تحسين النتائج لكل من المحللين.
Domain adaption in syntactic parsing is still a significant challenge. We address the issue of data imbalance between the in-domain and out-of-domain treebank typically used for the problem. We define domain adaptation as a Multi-task learning (MTL) problem, which allows us to train two parsers, one for each do-main. Our results show that the MTL approach is beneficial for the smaller treebank. For the larger treebank, we need to use loss weighting in order to avoid a decrease in performance be-low the single task. In order to determine towhat degree the data imbalance between two domains and the domain differences affect results, we also carry out an experiment with two imbalanced in-domain treebanks and show that loss weighting also improves performance in an in-domain setting. Given loss weighting in MTL, we can improve results for both parsers.
References used
https://aclanthology.org/
Large-scale multi-modal classification aim to distinguish between different multi-modal data, and it has drawn dramatically attentions since last decade. In this paper, we propose a multi-task learning-based framework for the multimodal classificatio
We propose a neural event coreference model in which event coreference is jointly trained with five tasks: trigger detection, entity coreference, anaphoricity determination, realis detection, and argument extraction. To guide the learning of this com
We present CoTexT, a pre-trained, transformer-based encoder-decoder model that learns the representative context between natural language (NL) and programming language (PL). Using self-supervision, CoTexT is pre-trained on large programming language
Aspect Category Sentiment Analysis (ACSA), which aims to identify fine-grained sentiment polarities of the aspect categories discussed in user reviews. ACSA is challenging and costly when conducting it into real-world applications, that mainly due to
Non-Autoregressive machine Translation (NAT) models have demonstrated significant inference speedup but suffer from inferior translation accuracy. The common practice to tackle the problem is transferring the Autoregressive machine Translation (AT) k