Thailand

Question answering on knowledge bases (KBQA) provides a user-friendly interface to the massive structured knowledge in KBs. It presents an array of intertwined challenges including non-i.i.d. generalization, large search space, and schema linking. Existing studies have tackled some of these challenges, but a unified solution that can address all the challenges is still lacking, which limits the applicability of KBQA on large-scale KBs. In this work, we propose a novel model, ArcaneQA, that aims to tackle all the aforementioned challenges and achieve strong (non-i.i.d.) generalization on large-scale KBs. Our model is based on two mutually boosting components: dynamic program induction, which dynamically prunes the search space, and dynamic contextualized encoding, which guides the search process on the large-scale KB via implicit schema linking. Experiment results on multiple popular KBQA datasets demonstrate the highly competitive performance of ArcaneQA in both effectiveness and efficiency, especially on non-i.i.d. generalization and complex questions involving multiple relations and functions.

IJCNLP-AACL 2021

ArcaneQA: Strongly Generalizable Question Answering on Large-Scale Knowledge Bases

non-i.i.d. generalization

knowledge base question answering

semantic parsing

**Welcome to ACL-IJCNLP 2021!**

The great event is jointly organized by the Association for Computational Linguistics (ACL) and Asian Federation of Natural Language Processing (AFNLP). 

As in previous years, the program of the conference includes a poster session, tutorials, workshops and demonstrations in addition to the main conference.


We were able to keep the registration fees similar to those charged for the virtual ACL 2020. The one fee allows attendance at the main conference and any/all tutorials and workshops. These fees would be $125 Regular Early and $175 Regular Late; $50 Student Early and $75 Student Late. Early registration closes at midnight July 11, 2021 (Eastern Daylight Time).

**Reminder:** It is ACL’s policy that at least one author of each accepted paper (including ACL Finding papers) must register for the conference.

**Reminder2:** Underline site will open closer to the event. If you already registered you will receive access detail

Registration is now open

The great event is jointly organized by the Association for Computational Linguistics (ACL) and Asian Federation of Natural Language Processing (AFNLP).

workshop paper

A common issue in real-world applications of named entity recognition and classification (NERC) is the absence of annotated data for the target entity classes during training. Zero-shot learning approaches address this issue by learning models from classes with training data that can predict classes without it. This paper presents the first approach for zero-shot NERC, introducing novel architectures that leverage the fact that textual descriptions for many entity classes occur naturally. We address the zero-shot NERC specific challenge that the not-an-entity class is not well defined as different entity classes are considered in training and testing.
For evaluation, we adapt two datasets, OntoNotes and MedMentions, emulating the difficulty of real-world zero-shot learning by testing models on the rarest entity classes. Our proposed approach outperforms baselines adapted from machine reading comprehension and zero-shot text classification. Furthermore, we assess the effect of different class descriptions for this task.

Leveraging Type Descriptions for Zero-shot Named Entity Recognition and Classification

Recently, word enhancement has become very popular for Chinese Named Entity Recognition (NER), reducing segmentation errors and increasing the semantic and boundary information of Chinese words. However, these methods tend to ignore the information of the Chinese character structure after integrating the lexical information. Chinese characters have evolved from pictographs since ancient times, and their structure often reflects more information about the characters. This paper presents a novel Multi-metadata Embedding based Cross-Transformer (MECT) to improve the performance of Chinese NER by fusing the structural information of Chinese characters. Specifically, we use multi-metadata embedding in a two-stream Transformer to integrate Chinese character features with the radical-level embedding. With the structural characteristics of Chinese characters, MECT can better capture the semantic information of Chinese characters for NER. The experimental results obtained on several well-known benchmarking datasets demonstrate the merits and superiority of the proposed MECT method.

MECT: Multi-Metadata Embedding based Cross-Transformer for Chinese Named Entity Recognition

We present an instance-based nearest neighbor approach to entity linking. In contrast to most prior entity retrieval systems which represent each entity with a single vector, we build a contextualized mention-encoder that learns to place similar mentions of the same entity closer in vector space than mentions of different entities. This approach allows all mentions of an entity to serve as ``class prototypes'' as inference involves retrieving from the full set of labeled entity mentions in the training set and applying the nearest mention neighbor's entity label. Our model is trained on a large multilingual corpus of mention pairs derived from Wikipedia hyperlinks, and performs nearest neighbor inference on an index of 700 million mentions. It is simpler to train, gives more interpretable predictions, and outperforms all other systems on two multilingual entity linking benchmarks.

MOLEMAN: Mention-Only Linking of Entities with a Mention Annotation Network

Models pre-trained on large-scale regular text corpora often do not work well for user-generated data where the language styles differ significantly from the mainstream text. Here we present Context-Aware Rule Injection (CARI), an innovative method for formality style transfer (FST) by injecting multiple rules into an end-to-end BERT-based encoder and decoder model. CARI is able to learn to select optimal rules based on context. The intrinsic evaluation showed that CARI achieved the new highest performance on the FST benchmark dataset. Our extrinsic evaluation showed that CARI can greatly improve the regular pre-trained models' performance on several tweet sentiment analysis tasks. Our contributions are as follows: 1.We propose a new method, CARI, to integrate rules for pre-trained language models. CARI is context-aware and can trained end-to-end with the downstream NLP applications. 2.We have achieved new state-of-the-art results for FST on the benchmark GYAFC dataset. 3.We are the first to evaluate FST methods with extrinsic evaluation and specifically on sentiment classification tasks. We show that CARI outperformed existing rule-based FST approaches for sentiment classification.

Improving Formality Style Transfer with Context-Aware Rule Injection

Exploring aspects of sentential meaning that are implicit or underspecified in context is important for sentence understanding. In this poster, we present our architecture based on mentions for revision requirements detection. The goal is to improve understandability, addressing some types of revisions, especially for the Replaced Pronoun type. We show that our mention-based system can predict replaced pronouns well on the mention-level. However, our combined sentence-level system does not improve on the sentence-level BERT baseline. We also present additional contrastive systems, and show results for each type of edit.

A Mention-Based System for Revision Requirements Detection

Some interpersonal verbs can implicitly attribute causality to either their subject or their object and are therefore said to carry an implicit causality (IC) bias. Through this bias, causal links can be inferred from a narrative, aiding language comprehension. We investigate whether pre-trained language models (PLMs) encode IC bias and use it at inference time. We find that to be the case, albeit to different degrees, for three distinct PLM architectures. However, causes do not always need to be implicit---when a cause is explicitly stated in a subordinate clause, an incongruent IC bias associated with the verb in the main clause leads to a delay in human processing. We hypothesize that the temporary challenge humans face in integrating the two contradicting signals, one from the lexical semantics of the verb, one from the sentence-level semantics, would be reflected in higher error rates for models on tasks dependent on causal links. The results of our study lend support to this hypothesis, suggesting that PLMs tend to prioritize lexical patterns over higher-order signals.

John praised Mary because he? Implicit Causality Bias and Its Interaction with Explicit Cues in LMs

CIs have been a long standing issue in theoreticallinguistics and the philosophy of language (Grice,1975; Karttunen and Peters, 1979; Bach, 1999;Potts, 2005, inter alia) and are highly frequent in natural, spontaneous language (Hautli-Janisz andButt, 2016). However, it is so far mostly scalar implicatures (Schuster et al., 2020, inter alia) and implicative verbs (and their veridicality) (Nairnet al., 2006; Ross and Pavlick, 2019, inter alia) that have drawn attention in theNLUcommunity. In this abstract we investigate a broader range ofCI items and constructions, with the ultimate goal of pro-viding a cross-linguistic dataset for modeling the interaction of explicit and implicit content in natural communication.

Variation in conventionally implicated content: An empirical study in English and German

Short-answer scoring is the task of assessing the correctness of a short text given as response to a question that can come from a variety of educational scenarios. 
As only content, not form, is important, the exact wording including the explicitness of an answer should not matter. However, many state-of-the-art scoring models heavily rely on lexical information, be it word embeddings in a neural network or n-grams in an SVM. Thus, the exact wording of an answer might very well make a difference. We therefore quantify to what extent implicit language phenomena occur in short answer datasets and examine the influence they have on automatic scoring performance. We find that the level of implicitness depends on the individual question, and that some phenomena are very frequent. Resolving implicit wording to explicit formulations indeed tends to improve automatic scoring performance.

Implicit Phenomena in Short-Answer Scoring Data

Commit message is a document that summarizes source code changes in natural language. A good commit message clearly shows the source code changes, so this enhances collaboration between developers. Therefore, our work is to develop a model that automatically writes the commit message.

To this end, we release 345K datasets consisting of code modification and commit messages in six programming languages (Python, PHP, Go, Java, JavaScript, and Ruby). Similar to the neural machine translation (NMT) model, using our dataset, we feed the code modification to the encoder input and the commit message to the decoder input and measure the result of the generated commit message with BLEU-4.

Also, we propose the following two training methods to improve the result of generating the commit message: (1) A method of preprocessing the input to feed the code modification to the encoder input. (2) A method that uses an initial weight suitable for the code domain to reduce the gap in contextual representation between programming language (PL) and natural language (NL).

CommitBERT: Commit Message Generation Using Pre-Trained Programming Language Model

Identifier names convey useful information aboutthe intended semantics of code. Name-based program analyses use this information, e.g., to detect bugs, to predict types,and to improve the readability of code. At the core of name-based analyses are semantic representations of identifiers, e.g., in the form of learned embeddings. The high-level goal ofsuch a representation is to encode whether two identifiers, e.g., len and size, are semantically similar. Unfortunately, it is currently unclear to what extent semantic representations matchthe semantic relatedness and similarity perceived by developers. This paper presents IdBench, the first benchmark for evaluating semantic representations against a ground truth created fromthousands of ratings by 500 software developers. We use IdBench to study state-of-the-art embedding techniques proposed for natural language, an embedding technique specifically designedfor source code, and lexical string distance functions. Our results show that the effectiveness of semantic representations varies significantly and that the best available embeddings successfully represent semantic relatedness. On the downside, no existing technique provides a satisfactory representation ofsemantic similarities, among other reasons because identifiers with opposing meanings are incorrectly considered to be similar,which may lead to fatal mistakes, e.g., in a refactoring tool.S tudying the strengths and weaknesses of the different techniques shows that they complement each other. As a first step toward exploiting this complementarity, we present an ensemble model that combines existing techniques and that clearly outperforms the best available semantic representation.

ArcaneQA: Strongly Generalizable Question Answering on Large-Scale Knowledge Bases

Next from IJCNLP-AACL 2021

Leveraging Type Descriptions for Zero-shot Named Entity Recognition and Classification

Similar lecture

Span-based Semantic Parsing for Compositional Generalization

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES