Semanlink - Europe AND Ville

- une première décrivant l'utilisation de données textuelles dans le cadre de la conception d'une opération marketing (définition de l'orientation à donner à l'image d'une marque). Les techniques NLP utilisées sont simples, mais leur utilisation comme outil pour aider les créatifs du marketing est originale.

- une deuxième dans le domaine juridique, très intéressante en termes de techniques mises en oeuvre. L'application vise à la valorisation de bases de contrats (un corpus de textes, sur lequel on souhaite effectuer des recherches complexes), en s'appuyant à la fois sur les techniques récentes de représentation de textes, et sur un knowledge graph (ontologie de termes juridiques). Pour ce qui est de le représentation de textes, ils utilisent Google BERT. Ce que permet BERT, c'est une forme de transfert d'apprentissage : BERT est un réseau de neurones profond entrainé de façon non supervisée, par et chez Google, sur une énorme quantité de textes, de façon à emmagasiner une connaissance sur une langue ("pre-trained language model"). Ces données (c'est à dire ce réseau pré-entrainé) sont mises à disposition par Google. Chacun peut ainsi affiner l'entrainement du réseau sur son propre corpus de textes, et ses propres données labellisées concernant le problème qu'il souhaite effectivement résoudre (par exemple, dans le cas juridique, la reconnaissance d'entités dans les contrats).

Le speaker rapporte des résultats sensiblement améliorés par rapport à ce qu'ils obtenaient auparavant en ce qui concerne la qualité de la représentation de phrases, et des problèmes de type classification de phrases ou reconnaissance d'entités (il note que la représentation de textes longs reste un problème ouvert). Les temps d'entrainement de BERT sur leur problème ne sont pas exorbitants (il parle de quelques heures de GPU, pas de jours ou de semaines de TPU comme pour l'entrainement initial).

- le dernier speaker quant à lui a présenté deux papiers de recherche, justement sur les techniques au cœur de BERT ("Transformer architecture").

2019-01-29 About

Romain Vial (Hyperlex) at Paris NLP meetup, slides

Tags:

2019-01-24 About

The memories around us – François Chollet – Medium

Tags:

2018-12-30 About

10 Exciting Ideas of 2018 in NLP

Tags:

2018-12-19 About

Monolingual data in NMT

Tags:

2018-12-04 About

Paris NLP Season 3 Meetup #2 – Paris NLP

Tags:

2018-12-04 About

Paris NLP - blog

Tags:

Paris NLP meetup

2018-11-29 About

Highlights of EMNLP 2018 – Chris Zhu – Medium

Tags:

EMNLP 2018

2018-11-25 About

Interesting Stuff at EMNLP (part II) – Valentin Malykh – Medium

Tags:

EMNLP 2018

2018-11-25 About

Interesting Stuff in EMNLP (part I) – Valentin Malykh – Medium

Tags:

EMNLP 2018

2018-11-25 About

Google AI Blog: Google at EMNLP 2018

Tags:

2018-11-25 About

Associative Multichannel Autoencoder for Multimodal Word Representation (2018)

Tags:

2018-11-15 About

EMNLP 2018 Thoughts and Notes · Supernatural Language Processing

Tags:

EMNLP 2018

2018-11-13 About

Trying to Understand Recurrent Neural Networks for Language Processing (slides)

Tags:

2018-11-11 About

Word Mover's Embedding: From Word2Vec to Document Embedding (2018)

Tags:

2018-11-10 About

EMNLP 2018 Highlights: Inductive bias, cross-lingual learning, and more

Tags:

2018-11-08 About

Adapting the Neural Encoder-Decoder Framework from Single to Multi-Document Summarization

Tags:

2018-11-06 About

Deep Chit-Chat: deep learning for chatbots (EMNLP 2018 Tutorial)

Tags:

by Dr Wei Wu (Microsoft Xiaolce - chatbot with 200 millions users in China) and Dr Rui Yan (Peking Univ)

- Chit-chat (casual, non goal oriented) open-domain. Must be relevant to the context and diverse (informative) to be engaging.
- why creating a chat? to prove an AI can speak like a human, commercial reasons, link to services.

Task oriented vs non task oriented: this tutorial is about the second one.

Retrieval based vs generation based.

Basic knowledge of DL for chatbots:

- word embeddings
- sentence embeddings (CNN, RNN)
- dialogue modeling: seq-to-seq with attention

Response selection for retrieval based chatbots:

- single turn response selection (slides 37-57)
    - framework 1: matching with seq embeddings
    - framework 2: matching with message-response interaction (46)
    - extension of 1: KnowledgeMatching with External Knowledge (53)
    - extension of 2: RepresentationsMatching with Multiple Levels of Representations (54)
    - insights from comparison between 1 and 2 (57)
- multi turn response selection (62)
    - context is now: mess + history
    - again, 2 frameworks

Emerging directions (79):

- matching with better representations
    - Self-Attention (82)
    - fusing multiple types of repr. But how to fuse matters (83)
    - pre-training

Learning a matching model for response selection (84)

Generation based models for chatbots:

- single turn generarion (89)
    - Basic generation model
        - seq2seq
        - Attention
        - Bi-directional modeling
- multi turn generation
    - Contexts are important
    - Context sensitive models
    - Hierarchical context modeling
    - Latent variable modeling
    - Hierarchical memory networks

Diversity in conversations (99)

Content introducing (106)

Additional elements (113)

- Topics in cnversation
- Emotions

Persona in chat:

- Persona
- ...
- Knowledge
- Common sense

RL and Adversarial learning in conversations (125)

Evaluation (132)

Future trends:

- Reasoning in dialogues
- X-grounded dialogues

2018-11-06 About

Joint Models in NLP - Slides - Tutorial (EMNLP 2018) - Yue Zhang

Tags:

**Joint models: solve 2 tasks at once.**

Related tasks: POS tagging, NER, chuncking. Pipeline tasks

Motivations:

- reduce error propagation
- information exchange between tasks

Challenges:

- Joint learning
- Search

History: statistical models. 2 kinds:

- Graph-Based Methods
    - Traditional solution:
        - Score each candidate, select the highest-scored output
        - Search-space typically exponential
- Transition-Based Methods
    - Transition-Based systems: Automata
        - State: partial result during decoding, Action: operations that can be applied for state transition
        - Output constructed incrementally

- Deep learning based model
    - Neural transition based models
    - Neural graph-based models
        - Cross task
            - Seminal work: Collobert, Ronan, et al. "Natural language processing (almost) from scratch."
            - Not all tasks are mutually beneficial
            - Ramachandran, et al.  “Unsupervised pretraining for sequence to sequence learning.”
            - Peters, Matthew E., et al. "Deep contextualized word representations." (ELMo)
            - "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding."
            - ULMFIT
            - Correlation between multi-task learning and pretraining
        - Cross lingual
        - Cross domain
        - Cross standard

2018-11-06 About

PROCEEDINGS of the BlackboxNLP Workshop

Tags:

BlackboxNLP (2018 workshop)

2018-11-06 About

Analyzing and interpreting neural networks for NLP (Workshop's Home page)

Tags:

BlackboxNLP (2018 workshop)

2018-11-06 About

Writing code for Natural language processing Research

Tags:

2018-11-05 About

Transfer learning with language models

Tags:

2018-11-05 About

Multi-Task Identification of Entities, Relations, and Coreference for Scientific Knowledge Graph Construction

Tags:

2018-11-04 About

Conference Schedule - EMNLP 2018

Tags:

EMNLP 2018

2018-11-04 About

Self-Governing Neural Networks for On-Device Short Text Classification - Sujith Ravi | Zornitsa Kozareva (2018)

Tags:

2018-11-02 About

EMNLP (2018) - ACL Anthology - Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing

Tags:

EMNLP 2018

2018-11-02 About

Deep Latent-Variable Models for Natural Language - Tutorial - harvardnlp

Tags:

2018-11-01 About

Facebook Research at EMNLP – Facebook Research

Tags:

2018-11-01 About

Trying to Understand Recurrent Neural Networks for Language Processing (tweets)

Tags:

2018-11-01 About

Writing Code for NLP Research, AllenNLP's tutorial at #emnlp2018

Tags:

2018-10-31 About

Tutorials - EMNLP 2018

Tags:

2018-10-31 About

Gilberto Gil and Caetano Veloso in London | Music | The Guardian

Tags:

2018-10-25 About

Paris NLP Season 3 Meetup #1 | Meetup

Tags:

2018-09-27 About

[1809.00782] Open Domain Question Answering Using Early Fusion of Knowledge Bases and Text

Tags:

2018-09-06 About

[1601.03764] Linear Algebraic Structure of Word Senses, with Applications to Polysemy

Tags:

> Here it is shown that multiple word senses reside
in linear superposition within the word
embedding and simple sparse coding can recover
vectors that approximately capture the
senses

> Each extracted word sense is accompanied by one of about  2000 “discourse atoms” that gives a succinct description of which other words co-occur with that word sense.

> The success of the approach is mathematically explained using a variant of
the random walk on discourses model

("random walk": a generative model for language). Under the assumptions of this model,  there
exists a linear relationship between the vector of a
word w and the vectors of the words in its contexts (It is not the average of the words in w's context, but in a given corpus the matrix of the linear relationship does not depend on w. It can be estimated, and so we can compute the embedding of a word from the contexts it belongs to)

[Related blog post](/doc/?uri=https%3A%2F%2Fwww.offconvex.org%2F2016%2F07%2F10%2Fembeddingspolysemy%2F)

2018-08-28 About

2018 Conference on Empirical Methods in Natural Language Processing - EMNLP 2018

Tags:

EMNLP 2018

2018-08-23 About

Paris NLP Meetup #6

Tags:

2018-07-26 About

Banksy peint les murs de Paris pour illustrer la crise des migrants

Tags:

2018-06-24 About

Module google/universal-sentence-encoder | TensorFlow

Tags:

2018-05-23 About

EC-WEB 2014 "Automotive ranges as e-commerce data"

Tags:

2017-04-12 About

Untergunther, réparateurs clandestins du patrimoine

Tags:

2017-02-04 About

Le regard de Robert McLiam Wilson sur les attentats du 13 novembre | GQ

Tags:

2016-06-12 About

Die Mörder sind unter uns (les assassins sont parmi nous)

Tags:

2016-05-25 About

A Year in Paris — An American Developer in Paris — Medium

Tags:

Paris

2016-05-07 About

Comment Jacques Jaujard a sauvé le Louvre

Tags:

2016-01-07 About

Bernard Stiegler : « Ce n’est qu’en projetant un véritable avenir qu’on pourra combattre Daech »

Tags:

2015-11-19 About

Attentats : le beau texte du « New York Times » était en fait un… commentaire

Tags:

2015-11-15 About

Eunice Barber, Paris 2003 Long Jump Women

Tags:

Championnats du monde à Paris-Saint Denis, 2003

2015-08-07 About

Les françaises championnes du monde du 4x100m (WC Paris 2003) HQ - YouTube

Tags:

Championnats du monde à Paris-Saint Denis, 2003

2015-08-07 About

Les français vice-champions du monde du 4x400m (WC Paris 2003) HQ - YouTube

Tags:

Championnats du monde à Paris-Saint Denis, 2003

2015-08-07 About

The Next 10 Years of Success

Tags:

2014-11-05 About

EC-WEB 2014 proceedings - Springer

Tags:

EC-Web'14

2014-09-13 About

Automotive range as e-commerce data (EC-WEB 2014)

Tags:

2014-08-04 About

EC-Web'14: The 15th International Conference on Electronic Commerce and Web Technologies

Tags:

EC-Web'14

2014-04-18 About

LOD2 Plenary and Open Data Meet-up in Mannheim

Tags:

2014-02-27 About

Semantic Web in Emergency Response Systems – UPDATE - semanticweb.com

Tags:

2013-10-18 About

Wikidata presentation at SemTechBiz Berlin 2012

Tags:

2013-07-10 About

Web Data Commons

Tags:

2013-07-06 About

Linked Data: Evolving the Web into a Global Data Space

Tags:

2012-11-22 About

Keynote Video and Updates from the Amsterdam Fire Department - semanticweb.com

Tags:

2012-11-08 About

Revue de presse SemWeb.Pro 2012 (SemWeb.Pro)

Tags:

SemWeb Pro 2012

2012-05-22 About

Descripteurs: De retour de SemWeb Pro 2012

Tags:

SemWeb Pro 2012

2012-05-15 About

SemWeb.Pro : le web sémantique démontre son utilité en entreprise #semwebpro

Tags:

2012-05-03 About

Gestion Documentaire (GED) Open Source par Nuxeo

Tags:

2012-05-02 About

L’empire romain et la société d’opulence énergétique : un parallèle | Oil Man

Tags:

2012-04-18 About

Linked Data et description du produit automobile (SemWeb.Pro)

Tags:

2012-02-13 About

KEYNOTE: Real-time Emergency Response Using Semantic Web Technology

Tags:

SemTechBiz Berlin 2012

2012-02-11 About

From Idea to Web - Creating Linked Data Apps

Tags:

2012-02-11 About

#SemTechBiz Berlin – Day 1 - semanticweb.com

Tags:

SemTechBiz Berlin 2012

2012-02-08 About

Semtech Berlin 2012 : un album sur Flickr

Tags:

SemTechBiz Berlin 2012

2012-02-08 About

SemTechBiz Berlin to Explore Semantics in the Auto Industry - semanticweb.com

Tags:

SemTechBiz Berlin 2012

2012-02-08 About

François-Paul Servant - semanticweb.com

Tags:

2012-02-08 About

#SemTechBiz Berlin – Day 2 - semanticweb.com

Tags:

2012-02-08 About

KEYNOTE Panel: Semantics in the Automotive Industry

Tags:

2011-12-17 About

SEMTECHBIZ Berlin 2012: The Semantic Tech & Business Conference

Tags:

SemTechBiz Berlin 2012

2011-11-02 About

Open data : l'exception culturelle? | Atelier Français

Tags:

2011-09-10 About

Rencontre avec Aldo Gangemi et Valentina Presutti - Les rencontres du Web de données (Paris) - Meetup

Tags:

2011-06-08 About

Linked Data - The Story So Far

Tags:

2011-03-08 About

Armadillo - Gestion documentaire multimédia

Tags:

2011-01-23 About

SemWeb.Pro 2011 : dossier de presse

Tags:

SemWeb Pro 2011

2011-01-21 About

Index of /semweblabs/semwebpro/rdfa

Tags:

2011-01-18 About

Compte-rendu du premier Meetup Web Sémantique « Blog Antidot

Tags:

2011-01-09 About

Nok sculpture, Louvre

Tags:

2011-01-04 About

SemWeb.Pro 2011 - January 17-18 Paris (SemWeb.Pro)

Tags:

2010-12-06 About

Comment réconcilier le SI legacy et le Web par le Web sémantique ? - Les rencontres du Web de données (Paris) - Meetup

Tags:

2010-12-01 About

blog.aksw.org » AKSW coordinates EU-funded research project LOD2 aiming to take the Web of Linked Data to the next level

Tags:

2010-09-06 About

Washington à la conquête du "9-3"

Tags:

2010-06-07 About

Au Danemark, la mutation d'Aalborg en ville écologique

Tags:

2010-05-28 About

4*400m, championnats du monde 1993, Paris

Tags:

2009-09-01 About

400m : remontée de Raquil aux championnats du monde St denis

Tags:

Championnats du monde à Paris-Saint Denis, 2003

2009-09-01 About

D2RQ — Lessons Learned

Tags:

D2RQ

2009-02-10 About

Relational database integration with RDF/OWL

Tags:

2009-02-07 About

Jena Assembler (D2RQ V0.5 - User Manual)

Tags:

2008-11-17 About

Relational database integration with RDF/OWL

Tags:

2008-06-04 About

ESWC2008 Conference Data

Tags:

2008-05-29 About

SemwebCampParis numéro 2 : Alexandre Passant

Tags:

2008-04-11 About

Developing Web 3.0 - JavaOne

Tags:

2008-04-08 About

Yahoo! Groupes : SemanticCampParis

Tags:

Semantic Camp Paris

2008-04-06 About

BarCamp wiki / SemanticCampParis

Tags:

2008-02-16 About

A qui la faute? de Victor Hugo

Tags:

2007-11-30 About

Victor Hugo : A qui la faute ?

Tags:

2007-09-26 About

[Linking-open-data] Returning to backlinks - Additional requirement: Paging

Tags:

2007-09-19 About

NG4J - Named Graphs API for Jena

Tags:

NG4J

2007-07-13 About

How to publish Linked Data on the Web?

Tags:

2007-07-12 About

How to publish Linked Data on the Web?

Tags:

2007-06-27 About

Pubby – A Linked Data Frontend for SPARQL Endpoints

Tags:

Pubby

2007-06-23 About

D2RQ - Treating Non-RDF Databases as Virtual RDF Graphs - Chris Bizer

Tags:

D2RQ

2007-06-23 About

[Linking-open-data] synonym URIs

Tags:

2007-06-13 About

A Aulnay, la victoire de M. Sarkozy vécue comme une "grosse claque"

Tags:

2007-05-08 About

Open Source Projects @ Freie Universität Berlin

Tags:

2007-04-20 About

Turning the Web into a Database

Tags:

2007-04-19 About

RDF Book Mashup

Tags:

2007-04-14 About

Disco - Hyperdata Browser

Tags:

Disco Hyperdata Browser

2007-04-03 About

Disco Hyperdata Browser

Tags:

2007-01-15 About

D2R Server publishing the DBLP Bibliography as Linked Data (Semantic Web grows 10%) from Chris Bizer

Tags:

2007-01-09 About

Semantic Web Client Library

Tags:

Semantic Web Client Library

2007-01-02 About

170, rue de Lourmel

Tags:

2006-09-23 About

170, rue de Lourmel

Tags:

2006-09-23 About

DEMO of the STITCH project ("Semantic Interoperability To access Cultural Heritage")

Tags:

The STITCH DEMO webserver demonstrates the added value of semantical integration of multiple collections. 
<br/>
The collections that are linked by the prototype RDF Multi-Faceted-Search browser are the Rijksmuseum Masterpieces Collection (ARIA) and the Medieval Illuminated Manuscripts at the National Library of the Netherlands<br/>
SKOS is used to enable mixing/merging of separate vocabularies:
Cultural-heritage collections are typically indexed with metadata derived from a range of different vocabularies, such as AAT, Iconclass and in-house standards. This presents a problem when one wants to use multiple collections in an interoperable way. In general, it is unrealistic to assume unification of vocabularies. Vocabularies have been developed in many sub-domains, each with their own emphasis and scope. Still, there is significant overlap between the vocabularies used for indexing. The prime research objective of this subproject is to develop theory, methods and tools for allowing metadata interoperability through semantic links between the vocabularies.

2006-04-07 About

Heredia - Les Trophées - La Trebbia : L'aube d'un jour sinistre a blanchi les hauteur

Tags:

2006-04-03 About

The 325 Project

Tags:

2006-01-30 About