45 research outputs found
HaritzPuerto/RCC: Final submission to RCC
No full text<p>Rich Context Competition. Team members: Giwon Hong, Minh-Son Cao and Haritz Puerto-San-Roman</p>
Let Me Know What to Ask: Interrogative-Word-Aware Question Generation
No full textQuestion Generation (QG) is a Natural Language Processing (NLP) task that aids advances in Question Answering (QA) and conversational assistants. Existing models focus on generating a question based on a text and possibly the answer to the generated question. They need to determine the type of interrogative word to be generated while having to pay attention to the grammar and vocabulary of the question. In this work, we propose Interrogative-Word-Aware Question Generation (IWAQG), a pipelined system composed of two modules: an interrogative word classifier and a QG model. The first module predicts the interrogative word that is provided to the second module to create the question. Owing to an increased recall of deciding the interrogative words to be used for the generated questions, the proposed model achieves new state-of-the-art results on the task of QG in SQuAD, improving from 46.58 to 47.69 in BLEU-1, 17.55 to 18.53 in BLEU-4, 21.24 to 22.33 in METEOR, and from 44.53 to 46.94 in ROUGE-L
Regularization of Distinct Strategies for Unsupervised Question Generation
No full textUnsupervised question answering (UQA) has been proposed to avoid the high cost of creating high-quality datasets for QA. One approach to UQA is to train a QA model with questions generated automatically. However, the generated questions are either too similar to a word sequence in the context or too drifted from the semantics of the context, thereby making it difficult to train a robust QA model. We propose a novel regularization method based on teacher-student architecture to avoid bias toward a particular question generation strategy and modulate the process of generating individual words when a question is generated. Our experiments demonstrate that we have achieved the goal of generating higher-quality questions for UQA across diverse QA datasets and tasks. We also show that this method can be useful for creating a QA model with few-shot learning
다중 홉 질의응답을 위한 의미역 기반 계층적 그래프 네트워크
No full text학위논문(석사) - 한국과학기술원 : 전산학부, 2021.2,[iv, 27 p. :]Multi-hop question answering requires the aggregation of information from several documents to find the answer to a question. Most prominent works approach this aggregation through entity graphs. However, they tend to overlook intra-sentence reasoning. In this work, we propose a graph structure whose main innovation is the use of semantic role labeling (SRL) arguments to explicitly model all the multi-hop reasoning steps, including the intra-sentence reasoning. Additionally, we propose a novel hierarchical graph2seq mechanism to fuse multi-hop and entity boundary information from the graph into the token embeddings of the context to enhance the answer span prediction task. We achieve competitive performance compared to the current state of the art and prove through extensive qualitative and quantitative experiments the effectiveness of SRL to model multi-hop reasoning, as well as the capabilities of our hierarchical graph2seq mechanism, which outperforms all previous approaches, to fuse graph information into the token embeddings.한국과학기술원 :전산학부
Topic Modeling for Analysing Similarity between Users in Twitter
No full textLa minería de datos en redes sociales está ganando importancia debido a que permite realizar campañas de marketing más precisas. Por ejemplo, Google realiza un análisis de todos nuestros datos: vídeos que vemos, términos que buscamos, páginas webs a las que accedemos, aplicaciones que descargamos, etc. para conocernos mejor y mostrarnos publicidad personalizada.
LDA es un modelo estadístico generativo para modelar documentos. Existen diversos algoritmos que dado un conjunto de documentos permiten obtener un modelo LDA que podría haber generado esos documentos. Con ese modelo es posible observar los temas usados en esos documentos y las palabras más relevantes para cada tema.
En el presente trabajo se pretende realizar una primera aproximación a la minería de datos en Twitter. Para ello, usando la API de Twitter se han descargado tweets de diversos usuarios y de sus seguidores. Posteriormente se han procesado esos Tweets generando documentos y se ha aplicado la implementación de Gensim del algoritmo Online LDA para obtener los temas de los documentos. Posteriormente, se han comparado los temas de los usuarios con los de sus seguidores.
También se proporciona un análisis del estado del arte de la minería de datos en Twitter
METHOD AND APPARATUS FOR OPERATING NEURAL NETWORK BASED ON SEMANTIC ROLE LABELING
No full text문장 내에 포함된 단어의 단어 임베딩 벡터를 생성하는 단계, 외부 도메인으로부터 수집된, 단어에 관한 계층적 단어 정보 및 단어 임베딩 벡터를 바탕으로 적어도 두 개의 맥락 어텐션 벡터를 생성하는 단계, 및 적어도 두 개의 맥락 어텐션 벡터를 결합하여 문장 임베딩 벡터를 생성하는 단계를 통해 문장을 인코딩하는 방법 및 인코더가 제공된다
METHOD AND APPARATUS FOR TRAINING UNSUPERVISED QUESTION GENERATION MODEL
No full text학습 장치의 동작 방법으로서, 문서와 정답에 대한 질의 생성 과정에서, 현재까지 추출된 단어 토큰들의 질의 타입을 판단하는 단계, 복수의 질의 생성 모델들 중에서, 판단한 질의 타입과 다른 타입의 특정 질의 생성 모델을, 다음 단어 토큰을 생성할 모델로 결정하는 단계, 상기 특정 질의 생성 모델이 입력 정보로부터, 어휘에 대해 예측한 확률 분포를 획득하는 단계, 그리고 상기 확률 분포를, 상기 입력 정보에 대한 정규화된 레이블로 생성하고, 상기 입력 정보와 상기 정규화된 레이블을 이용하여, 신규 질의 생성 모델을 학습시키는 단계를 포함한다
MetaQA: Combining Expert Agents for Multi-Skill Question Answering
No full textThe recent explosion of question answering (QA) datasets and models has
increased the interest in the generalization of models across multiple domains
and formats by either training on multiple datasets or by combining multiple
models. Despite the promising results of multi-dataset models, some domains or
QA formats may require specific architectures, and thus the adaptability of
these models might be limited. In addition, current approaches for combining
models disregard cues such as question-answer compatibility. In this work, we
propose to combine expert agents with a novel, flexible, and training-efficient
architecture that considers questions, answer predictions, and
answer-prediction confidence scores to select the best answer among a list of
answer candidates. Through quantitative and qualitative experiments we show
that our model i) creates a collaboration between agents that outperforms
previous multi-agent and multi-dataset approaches in both in-domain and
out-of-domain scenarios, ii) is highly data-efficient to train, and iii) can be
adapted to any QA format. We release our code and a dataset of answer
predictions from expert agents for 16 QA datasets to foster future developments
of multi-agent systems on https://github.com/UKPLab/MetaQA.Comment: Accepted at EACL 202
UKP-SQuARE: An Interactive Tool for Teaching Question Answering
No full textThe exponential growth of question answering (QA) has made it an indispensable topic in any Natural Language Processing (NLP) course. Additionally, the breadth of QA derived from this exponential growth makes it an ideal scenario for teaching related NLP topics such as information retrieval, explainability, and adversarial attacks among others. In this paper, we introduce UKP-SQuARE as a platform for QA education. This platform provides an interactive environment where students can run, compare, and analyze various QA models from different perspectives, such as general behavior, explainability, and robustness. Therefore, students can get a first-hand experience in different QA techniques during the class. Thanks to this, we propose a learner-centered approach for QA education in which students proactively learn theoretical concepts and acquire problem-solving skills through interactive exploration, experimentation, and practical assignments, rather than solely relying on traditional lectures. To evaluate the effectiveness of UKP-SQuARE in teaching scenarios, we adopted it in a postgraduate NLP course and surveyed the students after the course. Their positive feedback shows the platform’s effectiveness in their course and invites a wider adoption
UKP-SQuARE: An Interactive Tool for Teaching Question Answering
No full textThe exponential growth of question answering (QA) has made it an
indispensable topic in any Natural Language Processing (NLP) course.
Additionally, the breadth of QA derived from this exponential growth makes it
an ideal scenario for teaching related NLP topics such as information
retrieval, explainability, and adversarial attacks among others. In this paper,
we introduce UKP-SQuARE as a platform for QA education. This platform provides
an interactive environment where students can run, compare, and analyze various
QA models from different perspectives, such as general behavior,
explainability, and robustness. Therefore, students can get a first-hand
experience in different QA techniques during the class. Thanks to this, we
propose a learner-centered approach for QA education in which students
proactively learn theoretical concepts and acquire problem-solving skills
through interactive exploration, experimentation, and practical assignments,
rather than solely relying on traditional lectures. To evaluate the
effectiveness of UKP-SQuARE in teaching scenarios, we adopted it in a
postgraduate NLP course and surveyed the students after the course. Their
positive feedback shows the platform's effectiveness in their course and
invites a wider adoption.Comment: Accepted by BEA workshop, ACL202
