Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

"Text": models, code, and papers

Cross-Modal Learning via Pairwise Constraints

Nov 28, 2014
Ran He, Man Zhang, Liang Wang, Ye Ji, Qiyue Yin

In multimedia applications, the text and image components in a web document form a pairwise constraint that potentially indicates the same semantic concept. This paper studies cross-modal learning via the pairwise constraint, and aims to find the common structure hidden in different modalities. We first propose a compound regularization framework to deal with the pairwise constraint, which can be used as a general platform for developing cross-modal algorithms. For unsupervised learning, we propose a cross-modal subspace clustering method to learn a common structure for different modalities. For supervised learning, to reduce the semantic gap and the outliers in pairwise constraints, we propose a cross-modal matching method based on compound ?21 regularization along with an iteratively reweighted algorithm to find the global optimum. Extensive experiments demonstrate the benefits of joint text and image modeling with semantically induced pairwise constraints, and show that the proposed cross-modal methods can further reduce the semantic gap between different modalities and improve the clustering/retrieval accuracy.

* 12 pages, 5 figures, 70 references 

  Access Paper or Ask Questions

Software Infrastructure for Natural Language Processing

Feb 10, 1997
Hamish Cunningham, Kevin Humphreys, Robert Gaizauskas, Yorick Wilks

We classify and review current approaches to software infrastructure for research, development and delivery of NLP systems. The task is motivated by a discussion of current trends in the field of NLP and Language Engineering. We describe a system called GATE (a General Architecture for Text Engineering) that provides a software infrastructure on top of which heterogeneous NLP processing modules may be evaluated and refined individually, or may be combined into larger application systems. GATE aims to support both researchers and developers working on component technologies (e.g. parsing, tagging, morphological analysis) and those working on developing end-user applications (e.g. information extraction, text summarisation, document generation, machine translation, and second language learning). GATE promotes reuse of component technology, permits specialisation and collaboration in large-scale projects, and allows for the comparison and evaluation of alternative technologies. The first release of GATE is now available - see

* 5th Conference on Applied Natural Language Processing, 1997 
* LaTeX, uses aclap.sty, 8 pages 

  Access Paper or Ask Questions

WDV: A Broad Data Verbalisation Dataset Built from Wikidata

May 05, 2022
Gabriel Amaral, Odinaldo Rodrigues, Elena Simperl

Data verbalisation is a task of great importance in the current field of natural language processing, as there is great benefit in the transformation of our abundant structured and semi-structured data into human-readable formats. Verbalising Knowledge Graph (KG) data focuses on converting interconnected triple-based claims, formed of subject, predicate, and object, into text. Although KG verbalisation datasets exist for some KGs, there are still gaps in their fitness for use in many scenarios. This is especially true for Wikidata, where available datasets either loosely couple claim sets with textual information or heavily focus on predicates around biographies, cities, and countries. To address these gaps, we propose WDV, a large KG claim verbalisation dataset built from Wikidata, with a tight coupling between triples and text, covering a wide variety of entities and predicates. We also evaluate the quality of our verbalisations through a reusable workflow for measuring human-centred fluency and adequacy scores. Our data and code are openly available in the hopes of furthering research towards KG verbalisation.

  Access Paper or Ask Questions

How does fake news use a thumbnail? CLIP-based Multimodal Detection on the Unrepresentative News Image

Apr 27, 2022
Hyewon Choi, Yejun Yoon, Seunghyun Yoon, Kunwoo Park

This study investigates how fake news uses a thumbnail for a news article with a focus on whether a news article's thumbnail represents the news content correctly. A news article shared with an irrelevant thumbnail can mislead readers into having a wrong impression of the issue, especially in social media environments where users are less likely to click the link and consume the entire content. We propose to capture the degree of semantic incongruity in the multimodal relation by using the pretrained CLIP representation. From a source-level analysis, we found that fake news employs a more incongruous image to the main content than general news. Going further, we attempted to detect news articles with image-text incongruity. Evaluation experiments suggest that CLIP-based methods can successfully detect news articles in which the thumbnail is semantically irrelevant to news text. This study contributes to the research by providing a novel view on tackling online fake news and misinformation. Code and datasets are available at

* 9 pages, 8 figures including appendix figure, 2 tables. Published in Findings of ACL workshop, CONSTRAINT 2022 (Long paper). The manuscript is slightly revised after the camera ready version 

  Access Paper or Ask Questions

Conversational Question Answering on Heterogeneous Sources

Apr 25, 2022
Philipp Christmann, Rishiraj Saha Roy, Gerhard Weikum

Conversational question answering (ConvQA) tackles sequential information needs where contexts in follow-up questions are left implicit. Current ConvQA systems operate over homogeneous sources of information: either a knowledge base (KB), or a text corpus, or a collection of tables. This paper addresses the novel issue of jointly tapping into all of these together, this way boosting answer coverage and confidence. We present CONVINSE, an end-to-end pipeline for ConvQA over heterogeneous sources, operating in three stages: i) learning an explicit structured representation of an incoming question and its conversational context, ii) harnessing this frame-like representation to uniformly capture relevant evidences from KB, text, and tables, and iii) running a fusion-in-decoder model to generate the answer. We construct and release the first benchmark, ConvMix, for ConvQA over heterogeneous sources, comprising 3000 real-user conversations with 16000 questions, along with entity annotations, completed question utterances, and question paraphrases. Experiments demonstrate the viability and advantages of our method, compared to state-of-the-art baselines.

* SIGIR 2022 Research Track Long Paper 

  Access Paper or Ask Questions

A super-polynomial lower bound for learning nonparametric mixtures

Mar 28, 2022
Bryon Aragam, Wai Ming Tai

We study the problem of learning nonparametric distributions in a finite mixture, and establish a super-polynomial lower bound on the sample complexity of learning the component distributions in such models. Namely, we are given i.i.d. samples from $f$ where $$ f=\sum_{i=1}^k w_i f_i, \quad\sum_{i=1}^k w_i=1, \quad w_i>0 $$ and we are interested in learning each component $f_i$. Without any assumptions on $f_i$, this problem is ill-posed. In order to identify the components $f_i$, we assume that each $f_i$ can be written as a convolution of a Gaussian and a compactly supported density $\nu_i$ with $\text{supp}(\nu_i)\cap \text{supp}(\nu_j)=\emptyset$. Our main result shows that $\Omega((\frac{1}{\varepsilon})^{C\log\log \frac{1}{\varepsilon}})$ samples are required for estimating each $f_i$. The proof relies on a fast rate for approximation with Gaussians, which may be of independent interest. This result has important implications for the hardness of learning more general nonparametric latent variable models that arise in machine learning applications.

  Access Paper or Ask Questions

Contrastive Learning of Visual-Semantic Embeddings

Oct 17, 2021
Anurag Jain, Yashaswi Verma

Contrastive learning is a powerful technique to learn representations that are semantically distinctive and geometrically invariant. While most of the earlier approaches have demonstrated its effectiveness on single-modality learning tasks such as image classification, recently there have been a few attempts towards extending this idea to multi-modal data. In this paper, we propose two loss functions based on normalized cross-entropy to perform the task of learning joint visual-semantic embedding using batch contrastive training. In a batch, for a given anchor point from one modality, we consider its negatives only from another modality, and define our first contrastive loss based on expected violations incurred by all the negatives. Next, we update this loss and define the second contrastive loss based on the violation incurred only by the hardest negative. We compare our results with existing visual-semantic embedding methods on cross-modal image-to-text and text-to-image retrieval tasks using the MS-COCO and Flickr30K datasets, where we outperform the state-of-the-art on the MS-COCO dataset and achieve comparable results on the Flickr30K dataset.

  Access Paper or Ask Questions

Shaking Syntactic Trees on the Sesame Street: Multilingual Probing with Controllable Perturbations

Sep 28, 2021
Ekaterina Taktasheva, Vladislav Mikhailov, Ekaterina Artemova

Recent research has adopted a new experimental field centered around the concept of text perturbations which has revealed that shuffled word order has little to no impact on the downstream performance of Transformer-based language models across many NLP tasks. These findings contradict the common understanding of how the models encode hierarchical and structural information and even question if the word order is modeled with position embeddings. To this end, this paper proposes nine probing datasets organized by the type of \emph{controllable} text perturbation for three Indo-European languages with a varying degree of word order flexibility: English, Swedish and Russian. Based on the probing analysis of the M-BERT and M-BART models, we report that the syntactic sensitivity depends on the language and model pre-training objectives. We also find that the sensitivity grows across layers together with the increase of the perturbation granularity. Last but not least, we show that the models barely use the positional information to induce syntactic trees from their intermediate self-attention and contextualized representations.

* accepted to MRL @ EMNLP 2021 

  Access Paper or Ask Questions

Exploring Teacher-Student Learning Approach for Multi-lingual Speech-to-Intent Classification

Sep 28, 2021
Bidisha Sharma, Maulik Madhavi, Xuehao Zhou, Haizhou Li

End-to-end speech-to-intent classification has shown its advantage in harvesting information from both text and speech. In this paper, we study a technique to develop such an end-to-end system that supports multiple languages. To overcome the scarcity of multi-lingual speech corpus, we exploit knowledge from a pre-trained multi-lingual natural language processing model. Multi-lingual bidirectional encoder representations from transformers (mBERT) models are trained on multiple languages and hence expected to perform well in the multi-lingual scenario. In this work, we employ a teacher-student learning approach to sufficiently extract information from an mBERT model to train a multi-lingual speech model. In particular, we use synthesized speech generated from an English-Mandarin text corpus for analysis and training of a multi-lingual intent classification model. We also demonstrate that the teacher-student learning approach obtains an improved performance (91.02%) over the traditional end-to-end (89.40%) intent classification approach in a practical multi-lingual scenario.

  Access Paper or Ask Questions