News

Anahita will present on POS induction.
12 Aug 2019

In our lab meeting this week, Anahita will talk about part-of-speech induction. The title and abstract of her talk:

Inducing Multilingual Text Analysis Tools via Robust Projection across Aligned Corpora

Abstract: We will discuss the seminal work of Yarowsky and Ngai (2001) for inducing part-of-speech taggers for languages that have no annotated training data, but have translated text in a resource-rich language. This method does not assume any knowledge about the target language (no tagging dictionary is assumed), making it applicable to a wide array of resource-poor languages.

Tuesday, August 13thth, 12:00 p.m. TASC1 9408.

Zhenqi will present on entity recognition.
06 Aug 2019

In our lab meeting this week, Zhenqi will talk abou fine-grained entity recognition. The title and abstract of his talk:

Overview of latest works on fine-grained and ultra-fine entity recognition

Abstract: Recent works on fine-grained entity typing task has shift their attentions from encoding mention and context to a better type embedding. We will present four latest works on the fine-grained and ultra-fine entity recognition task and explore techniques such as zero-shot learning, graph convolution network and hyperbolic space of embedding entity types.

Tuesday, August 6th, 12:00 p.m. TASC1 9408.

Lindsey's practice talk.
23 Jul 2019

Lindsey will be practicing his thesis talk in this week’s lab meeting.

Understanding RNN States with Predictive Semantic Encodings and Adaptive Representations

Abstract: Recurrent neural networks are an effective and prevalent tool used to model sequential data such as natural language text. However, their deep nature and massive number of parameters pose a challenge for those intending to study precisely how they work. This is especially the case for researchers with the expertise to understand the mathematics behind these models at a macroscopic level, who often lack the tools to expose the microscopic details of what information they internally represent.

We present a combination of visual techniques to show some of the inner workings of recurrent neural networks and facilitate their study at a fine level of detail. Specifically, we introduce a consistent visual representation for vector data that is adaptive with respect to the available visual space. We tackle the problem of assigning meaning to hidden states by learning which outputs they produce and encoding this learned representation in a way that is quickly interpreted and relates to other elements of the visual design. These techniques are combined into a fully interactive visualization tool which is demonstrated to improve our understanding of common natural language processing tasks.

Tuesday, July 23rd, 12:00 p.m. TASC1 9408.

Jetic's practice talk.
16 Jul 2019

Jetic will be practicing his thesis presentation week’s lab meeting.

Exploiting Target Syntax with Structured Decoding

Abstract: *Incorporating structured target syntax in a neural machine translation model requires an effective tree-structured neural generator. We exploit a top-down tree-structured model called DRNN (Doubly-Recurrent Neural Networks) first proposed by Alvarez-Melis and Jaakola (2017) to create an NMT model called Seq2DRNN that combines a sequential encoder with tree-structured decoding augmented with a syntax-aware attention model. Unlike previous approaches to syntax-based NMT which use dependency parsing models our method uses constituency parsing which we argue provides useful information for translation. In addition, we use the syntactic structure of the sentence to add new connections to the tree-structured decoder neural network (Seq2DRNN+SynC). We compare our NMT model with sequential and state of the art syntax-based NMT models and show that our model produces more fluent translations with better reordering. Since our model is capable of doing translation and constituency parsing at the same time, we also compare our parsing accuracy against other neural parsing models. We also show that our proposed model is capable of learning even more strictly defined programming language syntax by modelling its Abstract Syntactic Trees, reaching new state-of-the-art exact match accuracy on Django dataset without having to resort to syntactic rule-based decoder. * Tuesday, July 16th, 12:00 p.m. TASC1 9408.

Zhenqi will present on entity recognition.
09 Jul 2019

In our lab meeting this week, Zhenqi will talk about some recent works on fine-grained entity recognition. The title and abstract of his talk:

Comparison among recent works on fine-grained entity recognition

Abstract: Fine-grained entity type classification (FETC) is the task of classifying an entity mention to a broad set of types. We will compare five papers in this area from the perspectives of feature extraction, model structure, loss function, hierarchical prediction and inference.

Tuesday, July 9th, 12:00 p.m. TASC1 9408.

Recent Publications