البشر قادرون على تعلم مفاهيم جديدة من أمثلة قليلة جدا؛ في المقابل، تحتاج خوارزميات التعلم في الآلة الحديثة عادة الآلاف من الأمثلة للقيام بذلك. في هذه الورقة، نقترح خوارزمية لتعلم مفاهيم جديدة من خلال تمثيلها كبرامج بشأن المفاهيم القائمة. وبهذه الطريقة، تعتبر مشكلة التعلم المفهوم بشكل طبيعي مشكلة تخليق برنامجا وتخصصت خوارزميةنا من بعض الأمثلة لتوليف برنامج يمثل مفهوم الرواية. بالإضافة إلى ذلك، نقوم بإجراء تحليل نظري لنهجنا للقضية التي يكون فيها البرنامج الذي يحدد مفهوم الرواية على تلك الموجودة خالية من السياق. نظهر أنه بالنظر إلى المحلل المحلل القائم على النحو المستفاد وقاعدة الإنتاج الجديدة، يمكننا زيادة المحلل بمحلل مع قاعدة الإنتاج بطريقة تعميم. نقيم نهجنا من خلال مفاهيم التعلم في مجال التحليل الدلالي الممتد إلى إعداد تعلم مفهوم الرواية القليلة، مما يظهر أن نهجنا يتفوق بشكل كبير على المحللين الدلالي العصبي المنتهي.
Humans are capable of learning novel concepts from very few examples; in contrast, state-of-the-art machine learning algorithms typically need thousands of examples to do so. In this paper, we propose an algorithm for learning novel concepts by representing them as programs over existing concepts. This way the concept learning problem is naturally a program synthesis problem and our algorithm learns from a few examples to synthesize a program representing the novel concept. In addition, we perform a theoretical analysis of our approach for the case where the program defining the novel concept over existing ones is context-free. We show that given a learned grammar-based parser and a novel production rule, we can augment the parser with the production rule in a way that provably generalizes. We evaluate our approach by learning concepts in the semantic parsing domain extended to the few-shot novel concept learning setting, showing that our approach significantly outperforms end-to-end neural semantic parsers.
References used
https://aclanthology.org/
Natural Language Processing (NLP) is increasingly relying on general end-to-end systems that need to handle many different linguistic phenomena and nuances. For example, a Natural Language Inference (NLI) system has to recognize sentiment, handle num
The ability to continuously expand knowledge over time and utilize it to rapidly generalize to new tasks is a key feature of human linguistic intelligence. Existing models that pursue rapid generalization to new tasks (e.g., few-shot learning methods
Semantic parsing aims at translating natural language (NL) utterances onto machine-interpretable programs, which can be executed against a real-world environment. The expensive annotation of utterance-program pairs has long been acknowledged as a maj
Synthesizing data for semantic parsing has gained increasing attention recently. However, most methods require handcrafted (high-precision) rules in their generative process, hindering the exploration of diverse unseen data. In this work, we propose
Meta-learning has recently been proposed to learn models and algorithms that can generalize from a handful of examples. However, applications to structured prediction and textual tasks pose challenges for meta-learning algorithms. In this paper, we a