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

Chrome logo Add to Chrome

Firefox logo Add to Firefox

"Topic": models, code, and papers

Hankel Matrix Nuclear Norm Regularized Tensor Completion for $N$-dimensional Exponential Signals

Mar 31, 2017
Jiaxi Ying, Hengfa Lu, Qingtao Wei, Jian-Feng Cai, Di Guo, Jihui Wu, Zhong Chen, Xiaobo Qu

Signals are generally modeled as a superposition of exponential functions in spectroscopy of chemistry, biology and medical imaging. For fast data acquisition or other inevitable reasons, however, only a small amount of samples may be acquired and thus how to recover the full signal becomes an active research topic. But existing approaches can not efficiently recover $N$-dimensional exponential signals with $N\geq 3$. In this paper, we study the problem of recovering N-dimensional (particularly $N\geq 3$) exponential signals from partial observations, and formulate this problem as a low-rank tensor completion problem with exponential factor vectors. The full signal is reconstructed by simultaneously exploiting the CANDECOMP/PARAFAC structure and the exponential structure of the associated factor vectors. The latter is promoted by minimizing an objective function involving the nuclear norm of Hankel matrices. Experimental results on simulated and real magnetic resonance spectroscopy data show that the proposed approach can successfully recover full signals from very limited samples and is robust to the estimated tensor rank.

* 15 pages, 12 figures 

  Access Paper or Ask Questions

Joint Event Detection and Entity Resolution: a Virtuous Cycle

Jul 18, 2016
Matthias Galle, Jean-Michel Renders, Guillaume Jacquet

Clustering web documents has numerous applications, such as aggregating news articles into meaningful events, detecting trends and hot topics on the Web, preserving diversity in search results, etc. At the same time, the importance of named entities and, in particular, the ability to recognize them and to solve the associated co-reference resolution problem are widely recognized as key enabling factors when mining, aggregating and comparing content on the Web. Instead of considering these two problems separately, we propose in this paper a method that tackles jointly the problem of clustering news articles into events and cross-document co-reference resolution of named entities. The co-occurrence of named entities in the same clusters is used as an additional signal to decide whether two referents should be merged into one entity. These refined entities can in turn be used as enhanced features to re-cluster the documents and then be refined again, entering into a virtuous cycle that improves simultaneously the performances of both tasks. We implemented a prototype system and report results using the TDT5 collection of news articles, demonstrating the potential of our approach.

  Access Paper or Ask Questions

NEMO: Extraction and normalization of organization names from PubMed affiliation strings

Jul 28, 2011
Siddhartha Jonnalagadda, Philip Topham

We propose NEMO, a system for extracting organization names in the affiliation and normalizing them to a canonical organization name. Our parsing process involves multi-layered rule matching with multiple dictionaries. The system achieves more than 98% f-score in extracting organization names. Our process of normalization that involves clustering based on local sequence alignment metrics and local learning based on finding connected components. A high precision was also observed in normalization. NEMO is the missing link in associating each biomedical paper and its authors to an organization name in its canonical form and the Geopolitical location of the organization. This research could potentially help in analyzing large social networks of organizations for landscaping a particular topic, improving performance of author disambiguation, adding weak links in the co-author network of authors, augmenting NLM's MARS system for correcting errors in OCR output of affiliation field, and automatically indexing the PubMed citations with the normalized organization name and country. Our system is available as a graphical user interface available for download along with this paper.

* Siddhartha Jonnalagadda, Philip Topham. NEMO: Extraction and normalization of organization names from PubMed affiliation strings. Journal of Biomedical Discovery and Collaboration, 2010 Oct 4;5:50-75 

  Access Paper or Ask Questions

Long-term Control for Dialogue Generation: Methods and Evaluation

May 15, 2022
Ramya Ramakrishnan, Hashan Buddhika Narangodage, Mauro Schilman, Kilian Q. Weinberger, Ryan McDonald

Current approaches for controlling dialogue response generation are primarily focused on high-level attributes like style, sentiment, or topic. In this work, we focus on constrained long-term dialogue generation, which involves more fine-grained control and requires a given set of control words to appear in generated responses. This setting requires a model to not only consider the generation of these control words in the immediate context, but also produce utterances that will encourage the generation of the words at some time in the (possibly distant) future. We define the problem of constrained long-term control for dialogue generation, identify gaps in current methods for evaluation, and propose new metrics that better measure long-term control. We also propose a retrieval-augmented method that improves performance of long-term controlled generation via logit modification techniques. We show through experiments on three task-oriented dialogue datasets that our metrics better assess dialogue control relative to current alternatives and that our method outperforms state-of-the-art constrained generation baselines.

  Access Paper or Ask Questions

Identifying Moments of Change from Longitudinal User Text

May 11, 2022
Adam Tsakalidis, Federico Nanni, Anthony Hills, Jenny Chim, Jiayu Song, Maria Liakata

Identifying changes in individuals' behaviour and mood, as observed via content shared on online platforms, is increasingly gaining importance. Most research to-date on this topic focuses on either: (a) identifying individuals at risk or with a certain mental health condition given a batch of posts or (b) providing equivalent labels at the post level. A disadvantage of such work is the lack of a strong temporal component and the inability to make longitudinal assessments following an individual's trajectory and allowing timely interventions. Here we define a new task, that of identifying moments of change in individuals on the basis of their shared content online. The changes we consider are sudden shifts in mood (switches) or gradual mood progression (escalations). We have created detailed guidelines for capturing moments of change and a corpus of 500 manually annotated user timelines (18.7K posts). We have developed a variety of baseline models drawing inspiration from related tasks and show that the best performance is obtained through context aware sequential modelling. We also introduce new metrics for capturing rare events in temporal windows.

* Preprint accepted for publication in ACL 2022 

  Access Paper or Ask Questions

Label Semantic Aware Pre-training for Few-shot Text Classification

Apr 14, 2022
Aaron Mueller, Jason Krone, Salvatore Romeo, Saab Mansour, Elman Mansimov, Yi Zhang, Dan Roth

In text classification tasks, useful information is encoded in the label names. Label semantic aware systems have leveraged this information for improved text classification performance during fine-tuning and prediction. However, use of label-semantics during pre-training has not been extensively explored. We therefore propose Label Semantic Aware Pre-training (LSAP) to improve the generalization and data efficiency of text classification systems. LSAP incorporates label semantics into pre-trained generative models (T5 in our case) by performing secondary pre-training on labeled sentences from a variety of domains. As domain-general pre-training requires large amounts of data, we develop a filtering and labeling pipeline to automatically create sentence-label pairs from unlabeled text. We perform experiments on intent (ATIS, Snips, TOPv2) and topic classification (AG News, Yahoo! Answers). LSAP obtains significant accuracy improvements over state-of-the-art models for few-shot text classification while maintaining performance comparable to state of the art in high-resource settings.

* Accepted at ACL 2022 

  Access Paper or Ask Questions

28 GHz Phased Array-Based Self-Interference Measurements for Millimeter Wave Full-Duplex

Mar 05, 2022
Aditya Chopra, Ian P. Roberts, Thomas Novlan, Jeffrey G. Andrews

We present measurements of the 28 GHz self-interference channel for full-duplex sectorized multi-panel millimeter wave (mmWave) systems, such as integrated access and backhaul. We measure the isolation between the input of a transmitting phased array panel and the output of a co-located receiving phased array panel, each of which is electronically steered across a number of directions in azimuth and elevation. In total, nearly 6.5 million measurements were taken in an anechoic chamber to densely inspect the directional nature of the coupling between 256-element phased arrays. We observe that highly directional mmWave beams do not necessarily offer widespread high isolation between transmitting and receiving arrays. Rather, our measurements indicate that steering the transmitter or receiver away from the other tends to offer higher isolation but even slight steering changes can lead to drastic variations in isolation. These measurements can be useful references when developing mmWave full-duplex solutions and can motivate a variety of future topics including beam/user selection and beamforming codebook design.

  Access Paper or Ask Questions

A Semi-Supervised Learning Approach with Two Teachers to Improve Breakdown Identification in Dialogues

Feb 22, 2022
Qian Lin, Hwee Tou Ng

Identifying breakdowns in ongoing dialogues helps to improve communication effectiveness. Most prior work on this topic relies on human annotated data and data augmentation to learn a classification model. While quality labeled dialogue data requires human annotation and is usually expensive to obtain, unlabeled data is easier to collect from various sources. In this paper, we propose a novel semi-supervised teacher-student learning framework to tackle this task. We introduce two teachers which are trained on labeled data and perturbed labeled data respectively. We leverage unlabeled data to improve classification in student training where we employ two teachers to refine the labeling of unlabeled data through teacher-student learning in a bootstrapping manner. Through our proposed training approach, the student can achieve improvements over single-teacher performance. Experimental results on the Dialogue Breakdown Detection Challenge dataset DBDC5 and Learning to Identify Follow-Up Questions dataset LIF show that our approach outperforms all previous published approaches as well as other supervised and semi-supervised baseline methods.

* Proceedings of the Thirty-Sixth AAAI Conference on Artificial Intelligence (2022) 
* 9 pages, 2 figures 

  Access Paper or Ask Questions

A Predictive Model for Student Performance in Classrooms Using Student Interactions With an eTextbook

Feb 16, 2022
Ahmed Abd Elrahman, Taysir Hassan A Soliman, Ahmed I. Taloba, Mohammed F. Farghally

With the rise of online eTextbooks and Massive Open Online Courses (MOOCs), a huge amount of data has been collected related to students' learning. With the careful analysis of this data, educators can gain useful insights into the performance of their students and their behavior in learning a particular topic. This paper proposes a new model for predicting student performance based on an analysis of how students interact with an interactive online eTextbook. By being able to predict students' performance early in the course, educators can easily identify students at risk and provide a suitable intervention. We considered two main issues the prediction of good/bad performance and the prediction of the final exam grade. To build the proposed model, we evaluated the most popular classification and regression algorithms on data from a data structures and algorithms course (CS2) offered in a large public research university. Random Forest Regression and Multiple Linear Regression have been applied in Regression. While Logistic Regression, decision tree, Random Forest Classifier, K Nearest Neighbors, and Support Vector Machine have been applied in classification.

* 21 pages,11 figures 

  Access Paper or Ask Questions

A Survey of Pretrained Language Models Based Text Generation

Feb 02, 2022
Junyi Li, Tianyi Tang, Wayne Xin Zhao, Jian-Yun Nie, Ji-Rong Wen

Text Generation aims to produce plausible and readable text in human language from input data. The resurgence of deep learning has greatly advanced this field by neural generation models, especially the paradigm of pretrained language models (PLMs). Grounding text generation on PLMs is seen as a promising direction in both academia and industry. In this survey, we present the recent advances achieved in the topic of PLMs for text generation. In detail, we begin with introducing three key points of applying PLMs to text generation: 1) how to encode the input data as representations preserving input semantics which can be fused into PLMs; 2) how to design a universal and performant architecture of PLMs served as generation models; and 3) how to optimize PLMs given the reference text and ensure the generated text satisfying special text properties. Then, we figure out several challenges and future directions within each key point. Next, we present a summary of various useful resources and typical text generation applications to work with PLMs. Finally, we conclude and summarize the contribution of this survey.

* 37 pages, 2 figures, 2 tables 

  Access Paper or Ask Questions