Recent large vision-language multimodal models pre-trained with huge amount of image-text pairs show remarkable performances in downstream tasks. However, the multimodal pre-training has limitations in terms of resources and training time when it comes to obtaining new models that surpass existing models. To overcome these issues, we propose TransferCVLM, a method of efficient knowledge transfer that integrates pre-trained uni-modal models (and cross-modal fusion-encoder) into a combined vision-language model (CVLM), without pre-training the CVLM with large amount of multimodal data, and then for each task application, fine-tunes the CVLM and transfers the multimodal knowledge of a teacher vision-language model to the CVLM by using knowledge distillation techniques. We demonstrate that 1) the fine-tuned CVLM performs comparable to other vision-language models of similar size, that 2) the multimodal knowledge transfer consistently enhances the CVLM, and the knowledge-transferred CVLM composed of large-size unimodal models outperforms the teacher multimodal model in most of downstream tasks, and that 3) TransferCVLM can also be used for model compression when using small-size unimodal models. We estimate that the training of TransferCVLM takes only 6% of pre-training of other vision-language models. Our code is available at https://github.com/DMCB-GIST/TransferCVLM.

Speculation detection is an important NLP task to identify text factuality. However, the extracted speculative information (e.g., speculative polarity, cue, and scope) lacks structure and poses challenges for direct utilization in downstream tasks. Open Information Extraction (OIE), on the other hand, extracts structured tuples as facts, without examining the certainty of these tuples. Bridging this gap between speculation detection and information extraction becomes imperative to generate structured speculative information and trustworthy relational tuples. Existing studies on speculation detection are defined at sentence level; but even if a sentence is determined to be speculative, not all factual tuples extracted from it are speculative. In this paper, we propose to study speculations in OIE tuples and determine whether a tuple is speculative. We formally define the research problem of tuple-level speculation detection. We then conduct detailed analysis on the LSOIE dataset which provides labels for speculative tuples. Lastly, we propose a baseline model SpecTup for this new research task.

Open Information Extraction (OIE) aims to extract relational tuples from open-domain sentences. Existing OIE systems split a sentence into tokens and recognize token spans as tuple relations and arguments. We instead propose Sentence as Chunk sequence (SaC) and recognize chunk spans as tuple relations and arguments. We argue that SaC has better properties for OIE than sentence as token sequence, and evaluate four choices of chunks (i.e., CoNLL chunks, OIA simple phrases, noun phrases, and spans from SpanOIE). Also, we propose a simple end-to-end BERT-based model, Chunk-OIE, for sentence chunking and tuple extraction on top of SaC. Chunk-OIE achieves state-of-the-art results on multiple OIE datasets, showing that SaC benefits the OIE task.

In this paper, we describe our system for ImageArg-2023 Shared Task that aims to identify an image’s stance towards a tweet and determine its persuasiveness score concerning a specific topic. In particular, the Shared Task proposes two subtasks viz. subtask (A) Multimodal Argument Stance (AS) Classification, and subtask (B) Multimodal Image Persuasiveness (IP) Classification, using a dataset composed of tweets (images and text) from controversial topics, namely gun control and abortion. For subtask A, we employ multiple transformer models using a text based approach to classify the argumentative stance of the tweet. For sub task B we adopted text based as well as multimodal learning methods to classify image persuasiveness of the tweet. Surprisingly, the text-based approach of the tweet overall performed better than the multimodal approaches considered. In summary, our best system achieved a F1 score of 0.85 for sub task (A) and 0.50 for subtask (B), and ranked 2nd in subtask (A) and 4th in subtask (B), among all teams submissions.

Prior research on dialogue state tracking (DST) is mostly based on written dialogue corpora. For spoken dialogues, the DST model trained on the written text should use the results (or hypothesis) of automatic speech recognition (ASR) as input. But ASR hypothesis often includes errors, which leads to significant performance drop for spoken dialogue state tracking. We address the issue by developing the following ASR error correction modules. First, we train a model to convert ASR hypothesis to ground truth user utterance, which can fix frequent patterns of errors. The model takes ASR hypotheses of two ASR models as input and fine-tuned in two stages. The corrected hypothesis is fed into a large scale pre-trained encoder-decoder model (T5) for DST training and inference. Second, if an output slot value from the encoder-decoder model is a name, we compare it with names in a dictionary crawled from Web sites and, if feasible, replace with the crawled name of the shortest edit distance. Third, we fix errors of temporal expressions in ASR hypothesis by using hand-crafted rules. Experiment results on the DSTC 11 speech-aware dataset, which is built on the popular MultiWOZ task (version 2.1), show that our proposed method can effectively mitigate the performance drop when moving from written text to spoken conversations.

Open Information Extraction (OpenIE) aims to extract relational tuples from open-domain sentences. Traditional rule-based or statistical models were developed based on syntactic structure of sentence, identified by syntactic parsers. However, previous neural OpenIE models under-explored the useful syntactic information. In this paper, we model both constituency and dependency trees into word-level graphs, and enable neural OpenIE to learn from the syntactic structures. To better fuse heterogeneous information from the two graphs, we adopt multi-view learning to capture multiple relationships from them. Finally, the finetuned constituency and dependency representations are aggregated with sentential semantic representations for tuple generation. Experiments show that both constituency and dependency information, and the multi-view learning are effective.

We present a multi-modal deep learning system for the Multimedia Automatic Misogyny Identification (MAMI) challenge, a SemEval task of identifying and classifying misogynistic messages in online memes. We adapt multi-task learning for the multimodal subtasks of the MAMI challenge to transfer knowledge among the correlated subtasks. We also leverage on ensemble learning for synergistic integration of models individually trained for the subtasks. We finally discuss about errors of the system to provide useful insights for future work.

Complex question answering over knowledge base remains as a challenging task because it involves reasoning over multiple pieces of information, including intermediate entities/relations and other constraints. Previous methods simplify the SPARQL query of a question into such forms as a list or a graph, missing such constraints as “filter” and “order_by”, and present models specialized for generating those simplified forms from a given question. We instead introduce a novel approach that directly generates an executable SPARQL query without simplification, addressing the issue of generating unseen entities. We adapt large scale pre-trained encoder-decoder models and show that our method significantly outperforms the previous methods and also that our method has higher interpretability and computational efficiency than the previous methods.