Convolutional Neural Networks vs. Convolution Kernels: Feature Engineering for Answer Sentence Reranking
By Kateryna Tymoshenko, Daniele Bonadiman, Alessandro Moschitti. 2016
62 citations recorded by the source corpus.
In this paper, we study, compare and combine two state-of-the-art approaches to automatic feature engineering: Convolution Tree Kernels (CTKs) and Convolutional Neural Networks (CNNs) for learning to rank answer sentences in a Question Answering (QA) setting. When dealing with QA, the key aspect is to encode relational information between the constituents of question and answer in learning algorithms. For this purpose, we propose novel CNNs using relational information and combined them with relational CTKs. The results show that (i) both approaches achieve the state of the art on a question answering task, where CTKs produce higher accuracy and (ii) combining such methods leads to unprecedented high results.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Natural Language Processing
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Mohit Bansal
- Luke S. Zettlemoyer
- Jianfeng Gao
- Percy Liang
- Yonatan Belinkov
- Noah A. Smith
- Hannaneh Hajishirzi
- Graham Neubig
In Computer Vision and Pattern Recognition
- Kenton C. T. Lee
- Ankur P. Parikh
- Yonatan Belinkov
- Nan Duan
- James R. Glass
- Tom Kwiatkowski
- Preslav Nakov
- Hamdy Mubarak
In Natural Language Processing
- Luke S. Zettlemoyer
- Yonatan Belinkov
- Mohit Bansal
- Percy Liang
- Dipanjan Das
- Nan Duan
- Noah A. Smith
- M. Zhou
Citations in the corpus
- Learning Recurrent Span Representations for Extractive Question Answering (2016, background): “Machine learned extractive question answering systems, such as the one presented here, learn a predictor function f ( q , p ) → a from a training dataset of (cid:104) q , p , a (cid:105) triples.”
- Learning What is Essential in Questions (2017, neutral): “For instance, Neural Networks based models (Wang et al., 2016; Tymoshenko et al., 2016; Yin et al., 2016) implicitly compute some kind of attention.”
- Question Generation for Question Answering (2017, methodology): “7 Question Generation for QA A typical answer sentence selection method, such as (Yin et al., 2016; Santos et al., 2016; Miller et al., 2016; Tymoshenko et al., 2016), computes the relevance score between input question Q and each answer candidate A , and selects the one with the highest relevance…”
- SemEval-2016 Task 3: Community Question Answering (2019, background): “It should be noted that ConvKN and KeLP used tree kernels with relational links (Tymoshenko and Moschitti, 2015; Tymoshenko et al., 2016), i.e., the questions are aligned with the comments (or with the other questions) by means of a special REL tag, directly annotated in the parse trees.”