Language Tech Market News
52.7K views | +51 today
Follow
Language Tech Market News
The Home of Multilingual Intelligence
Curated by LT-Innovate
Your new post is loading...
Your new post is loading...
Scooped by LT-Innovate
Scoop.it!

A Review of the Neural History of Natural Language Processing

A Review of the Neural History of Natural Language Processing | Language Tech Market News | Scoop.it

S; Ruder: This is the first blog post in a two-part series. The series expands on the Frontiers of Natural Language Processing session organized by Herman Kamper and me at the Deep Learning Indaba 2018. Slides of the entire session can be found here. This post will discuss major recent advances in NLP focusing on neural network-based methods. The second post will discuss open problems in NLP.

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

How Research Should Improve Current #NLP’s Generalization Problem

How Research Should Improve Current #NLP’s Generalization Problem | Language Tech Market News | Scoop.it
  • We should use more inductive biases, but we have to work out what are the most suitable ways to integrate them into neural architectures such that they really lead to expected improvements.
  • We have to enhance pattern-matching state-of-the-art models with some notion of human-like common sense that will enable them to capture the higher-order relationships among facts, entities, events or activities. But mining common sense is challenging, so we need new, creative ways of extracting common sense.
  • Finally, we should deal with unseen distributions and unseen tasks, otherwise “any expressive model with enough data will do the job.” Obviously, training such models is harder and results will not immediately be impressive. As researchers we have to be bold with developing such models, and as reviewers we should not penalize work that tries to do so.
more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Handy Research Resource: Tracking Progress in #NLProc

Sebastian Ruder's doc tracks the progress in Natural Language Processing (NLP) and give an overview of the state-of-the-art across the most common NLP tasks and their corresponding datasets. It covers traditional and core NLP tasks such as dependency parsing and part-of-speech tagging as well as more recent ones such as reading comprehension and natural language inference. The main objective is to provide the reader with a quick overview of benchmark datasets and the state-of-the-art for their task of interest, which serves as a stepping stone for further research. To this end, if there is a place where results for a task are already published and regularly maintained, such as a public leaderboard, the reader will be pointed there.

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

New Models to Bring Low-resource Languages and Spoken Dialects into #NeuralMT

New Models to Bring Low-resource Languages and Spoken Dialects into #NeuralMT | Language Tech Market News | Scoop.it

Microsoft Research. We tackled the challenge of insufficient parallel data using a Semi-Supervised Universal Neural Machine Translation approach that requires only a few thousand parallel sentences for an extremely low-resource language to achieve a high-quality machine translation system. This exciting research will be presented at the 16th Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-2018).

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

@Clearstream Partnership to Explore #AI and #NLP in Process Automation

@Clearstream Partnership to Explore #AI and #NLP in Process Automation | Language Tech Market News | Scoop.it

Clearstream, part of Deutsche Börse Group, has entered a collaborative research partnership with the Interdisciplinary Centre for Security, Reliability and Trust (SnT) at the University of Luxembourg. Over the next four years, a joint research project will explore the potential of Artificial Intelligence (AI) and Natural Language Processing (NLP) in process automation in order to simplify and standardize requirements analysis. The project aims to eventually render processes more cost- and time-efficient.

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

NarrativeQA: The Difficulty of Getting Machines to Comprehend Books & Movie Scripts

To encourage progress on deeper comprehension of language, we present a new dataset and set of tasks in which the reader must answer questions about stories by reading entire books or movie scripts. These tasks are designed so that successfully answering their questions requires understanding the underlying narrative rather than relying on shallow pattern matching or salience. We show that although humans solve the tasks easily, standard RC models struggle on the tasks presented here. We provide an analysis of the dataset and the challenges it presents.

LT-Innovate's insight:

New Deep Mind paper.

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Getting Closer to Predicting Human Speech by Studying Bird-song

Getting Closer to Predicting Human Speech by Studying Bird-song | Language Tech Market News | Scoop.it

Gentner and his team hope their finches will help make it possible. “We have demonstrated a [brain-machine interface] for a complex communication signal, using an animal model for human speech,” they write. They add that “our approach also provides a valuable proving ground for biomedical speech-prosthetic devices.”
In other words, we’re a little closer to texting from our brains.

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Help Measure AI Progress in Research

Help Measure AI Progress in Research | Language Tech Market News | Scoop.it
This pilot project collects problems and metrics/datasets from the AI research literature, and tracks progress on them. You can use this notebook to see how things are progressing in specific subfields or AI/ML as a whole, as a place to report new results you've obtained, as a place to look for problems that might benefit from having new datasets/metrics designed for them, or as a source to build on for data science projects. At EFF, we're ultimately most interested in how this data can influence our understanding of the likely implications of AI. To begin with, we're focused on gathering it.
LT-Innovate's insight:

Useful collection of informaion on AI research, driven by data-sharing ethos.

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Can Global Semantic Context Improve Neural Language Models? from @Apple's Machine Learning Journal

Can Global Semantic Context Improve Neural Language Models? from @Apple's Machine Learning Journal | Language Tech Market News | Scoop.it

From Apple's Machine Learning Journal Vol. 1, Issue 11 ∙ September 2018
by Frameworks Natural Language Processing Team
Entering text on your iPhone, discovering news articles you might enjoy, finding out answers to questions you may have, and many other language-related tasks depend upon robust natural language processing (NLP) models. Word embeddings are a category of NLP models that mathematically map words to numerical vectors. This capability makes it fairly straightforward to find numerically similar vectors or vector clusters, then reverse the mapping to get relevant linguistic information. Such models are at the heart of familiar apps like News, search, Siri, keyboards, and Maps.

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Using #NMT to Ensure Low-Latency Speech Translation

Using #NMT to Ensure Low-Latency Speech Translation | Language Tech Market News | Scoop.it

Researchers from the Karlsruhe Institute of Technology (KIT), in Germany, "In this work, we aim to remedy the problem of partial sentence translation in NMT," the researchers wrote. "Ideally, we want a model that is able to generate appropriate translations for incomplete sentences, without any compromise during other translation use cases."  Their adaptation of NMT achieved high-quality translations at low latency, minimizing the number of corrected words by 45 percent. In the future, their study could have meaningful practical implications, helping to develop better tools for real-time speech translation.

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Google Plans to Build AI Research Center in Ghana

Google Plans to Build AI Research Center in Ghana | Language Tech Market News | Scoop.it

Google plans to open an artificial-intelligence research center in Accra, Ghana, the latest in a string of investments the tech company has made in Africa. The research center will focus on using AI in areas such as healthcare, agriculture and education, Google said. “We’re committed to collaborating with local universities and research centers, as well as working with policy makers on the potential uses of AI in Africa,” the company In a blog post on Wednesday.

LT-Innovate's insight:

Hope Google does some data collection in Nigeria too - for languages such as Hausa, Igbo, Yoruba, Urhobo, Ibibio, Edo, Fulfulde and Kanuri. And note that Facebook's opened a research centre in Lagos.

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

The Birth of Collective Memories: Analyzing Emerging Entities in Text Streams

The Birth of Collective Memories: Analyzing Emerging Entities in Text Streams | Language Tech Market News | Scoop.it
The online text streams we use for our analysis comprise of social media and news streams, and span over 579 million documents in a time span of 18 months. We discover two main emergence patterns: entities that emerge in a “bursty” fashion, that is, that appear in public discourse without a precedent, blast into activity and transition into collective memory. Other entities display a “delayed” pattern, where they appear in public discourse, experience a period of inactivity, and then resurface before transitioning into our cultural collective memory
LT-Innovate's insight:

Before topics get institutionalised in Wikipedia. Strange stuff. 

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Helping AI To Learn The Flow Of Conversation

Helping AI To Learn The Flow Of Conversation | Language Tech Market News | Scoop.it

Technical paper: A recent paper from Osaka University aims to make AI-based systems better at lexical acquisition.  Their method utilizes implicit confirmation to allow a system to understand the category of a word across multiple exchanges.  It does this by confirming whether or not its predictions are true in real-time as the conversation flows.

LT-Innovate's insight:

Starting on the next stage beyond scripted alternatives in AI-driven conversations

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Video: @AlexWaibel Reviews His Career in Language Technology R&D and Entrepreneurship

Video: @AlexWaibel Reviews His Career in Language Technology R&D and Entrepreneurship | Language Tech Market News | Scoop.it
“The problem today, is no longer how to connect to people, but how to deal with 6,000 languages”. Alex Waibel – Professor at Carnegie Mellon & Karlsruhe Institute of Technology and Member of the Advisory Board at Pi School, School of AI – opened the day with his amazing presentation, comparing human interactions with the latest progresses on technology. “Is it possible to build dialogue systems, so I can speak in one language and the other person can hear me and respond on its own language?”, he asks
more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Alibaba to Invest $15B in Tech, with Research Labs Around the World

Alibaba to Invest $15B in Tech, with Research Labs Around the World | Language Tech Market News | Scoop.it

Alibaba announced it was launching a new research institute dubbed the “Academy for Discovery, Adventure, Momentum, and Outlook” (DAMO), which includes seven new research labs in China, the United States, Russia, Israel, and Singapore.
In addition to the new research academy, Alibaba is also committing US$15 billion to research and development over the next three years – almost doubling its average annual spend. For the company’s 2017 financial year ending in March, the company spent about US$2.6 billion on product development.

LT-Innovate's insight:

R&D moving up the value chain?

more...
No comment yet.
Scooped by LT-Innovate
Scoop.it!

Translation Research Gets Boost from Ireland’s €50M ADAPT Investment

Ireland’s Minister for Skills, Research, and Innovation, has announced a total of EUR 50 million (USD 54 million) of fresh funding for ADAPT, the research center of Science Foundation Ireland (SFI). The investment intends to position Ireland as a leader in next generation digital content technologies, including MT and other translation projects.The money is split between EUR 24 million coming from SFI and a further EUR 26 million coming from 19 different industry partners. 

more...
No comment yet.