There have been various types of pretraining architectures including autoregressive models (e.g., GPT), autoencoding models (e.g., BERT), and encoder-decoder models (e.g., T5). On the other hand, NLP tasks are different in nature, with three main cat
egories being classification, unconditional generation, and conditional generation. However, none of the pretraining frameworks performs the best for all tasks, which introduces inconvenience for model development and selection. We propose a novel pretraining framework GLM (General Language Model) to address this challenge. Compared to previous work, our architecture has three major benefits: (1) it performs well on classification, unconditional generation, and conditional generation tasks with one single pretrained model; (2) it outperforms BERT-like models on classification due to improved pretrain-finetune consistency; (3) it naturally handles variable-length blank filling which is crucial for many downstream tasks. Empirically, GLM substantially outperforms BERT on the SuperGLUE natural language understanding benchmark with the same amount of pre-training data. Moreover, GLM with 1.25x parameters of BERT-Large achieves the best performance in NLU, conditional and unconditional generation at the same time, which demonstrates its generalizability to different downstream tasks.
Most modern Information Extraction (IE) systems are implemented as sequential taggers and only model local dependencies. Non-local and non-sequential context is, however, a valuable source of information to improve predictions. In this paper, we intr
oduce GraphIE, a framework that operates over a graph representing a broad set of dependencies between textual units (i.e. words or sentences). The algorithm propagates information between connected nodes through graph convolutions, generating a richer representation that can be exploited to improve word-level predictions. Evaluation on three different tasks --- namely textual, social media and visual information extraction --- shows that GraphIE consistently outperforms the state-of-the-art sequence tagging model by a significant margin.
We study the extent to which we can infer users geographical locations from social media. Location inference from social media can benefit many applications, such as disaster management, targeted advertising, and news content tailoring. The challenge
s, however, lie in the limited amount of labeled data and the large scale of social networks. In this paper, we formalize the problem of inferring location from social media into a semi-supervised factor graph model (SSFGM). The model provides a probabilistic framework in which various sources of information (e.g., content and social network) can be combined together. We design a two-layer neural network to learn feature representations, and incorporate the learned latent features into SSFGM. To deal with the large-scale problem, we propose a Two-Chain Sampling (TCS) algorithm to learn SSFGM. The algorithm achieves a good trade-off between accuracy and efficiency. Experiments on Twitter and Weibo show that the proposed TCS algorithm for SSFGM can substantially improve the inference accuracy over several state-of-the-art methods. More importantly, TCS achieves over 100x speedup comparing with traditional propagation-based methods (e.g., loopy belief propagation).
Measuring research impact and ranking academic achievement are important and challenging problems. Having an objective picture of research institution is particularly valuable for students, parents and funding agencies, and also attracts attention fr
om government and industry. KDD Cup 2016 proposes the paper acceptance rank prediction task, in which the participants are asked to rank the importance of institutions based on predicting how many of their papers will be accepted at the 8 top conferences in computer science. In our work, we adopt a three-step feature engineering method, including basic features definition, finding similar conferences to enhance the feature set, and dimension reduction using PCA. We propose three ranking models and the ensemble methods for combining such models. Our experiment verifies the effectiveness of our approach. In KDD Cup 2016, we achieved the overall rank of the 2nd place.
In online question-and-answer (QA) websites like Quora, one central issue is to find (invite) users who are able to provide answers to a given question and at the same time would be unlikely to say no to the invitation. The challenge is how to trade
off the matching degree between users expertise and the question topic, and the likelihood of positive response from the invited users. In this paper, we formally formulate the problem and develop a weakly supervised factor graph (WeakFG) model to address the problem. The model explicitly captures expertise matching degree between questions and users. To model the likelihood that an invited user is willing to answer a specific question, we incorporate a set of correlations based on social identity theory into the WeakFG model. We use two different genres of datasets: QA-Expert and Paper-Reviewer, to validate the proposed model. Our experimental results show that the proposed model can significantly outperform (+1.5-10.7% by MAP) the state-of-the-art algorithms for matching users (experts) with community questions. We have also developed an online system to further demonstrate the advantages of the proposed method.
