Semanlink - Favoris ; AI girls and guys AND Good

[2305.12517] Retrieving Texts based on Abstract Descriptions

Tags:

2023-06-15 About

Building Scalable, Explainable, and Adaptive NLP Models with Retrieval | SAIL Blog

Tags:

2021-10-07 About

Unsupervised text classification with word embeddings - Max Halford

Tags:

2020-10-05 About

[1909.01259] Neural Attentive Bag-of-Entities Model for Text Classification

Tags:

A model that performs **text classification using entities in a knowledge base**.

> Entities provide unambiguous and relevant semantic signals that are beneficial for capturing semantics in texts. We combine **simple high-recall entity detection based on a dictionary** (word->list of entities), to detect entities in a document, with a novel neural **attention mechanism that enables the model to focus on a small number of unambiguous and relevant entities**.

2 steps:

1. Entity detection
2. Classification using the detected entities (+text) as inputs

Regarding entity linking, a local model which uses cosine
similarity between the embedding of the target
entity and the word-based representation of
the document to capture the relevance of an entity
given a document.

Embeddings from the KB: computed using [#Wikipedia2Vec](tag:wikipedia2vec) (similar words and entities
close to one another in a unified vector space)

Model using attention, with 2 features :

- cosine similarity between the
embedding of the entity and the word based
representation of the document
- the probability that the entity
name refers to the entity in KB.

Somewhat [related](doc:2020/01/investigating_entity_knowledge_)

### Conclusion:

>a neural
network model that performs text classification using
entities in Wikipedia. We combined simple
dictionary-based entity detection with a neural attention
mechanism to enable the model to focus
on a small number of unambiguous and relevant
entities in a document.

2020-09-02 About

[1807.00082] Amanuensis: The Programmer's Apprentice

Tags:

**The use of natural language to facilitate communication
between the expert programmer and apprentice AI system.**

> an overview of the material covered in a course taught at Stanford in the spring quarter of 2018. The course draws upon **insight from cognitive and systems neuroscience to implement hybrid connectionist and symbolic reasoning systems** that leverage and extend the state of the art in machine learning **by integrating human and machine intelligence**. As a concrete example we focus on digital assistants that learn from continuous dialog with an expert software engineer while providing initial value as powerful analytical, computational and mathematical savants.

> [#Dehaene](/tag/stanislas_dehaene)'s work extends the [#Global Workspace Theory](/tag/global_workspace_theory) of Bernard Baars. Dehaene’s version of the theory combined with Yoshua Bengio’s concept of a [#consciousness prior](/tag/consciousness_prior.html) and deep reinforcement learning suggest a model for constructing and maintaining the cognitive states that arise and persist during complex problem solving.

2019-11-12 About

The knowledge graph as the default data model for learning on heterogeneous knowledge (2017)

Tags:

2019-08-22 About

Transformers from scratch | Peter Bloem

Tags:

2019-08-21 About

Jeremy Howard on Twitter: "Such a ridiculously simple idea couldn't possibly work, could it? Or... could it? "

Tags:

2019-02-09 About

Learning Deep Architectures for AI By Yoshua Bengio (2009)

Tags:

2018-11-06 About

A Review of the Recent History of Natural Language Processing - AYLIEN

Tags:

2018-10-02 About

Simple and efficient semantic embeddings for rare words, n-grams, and language features – Off the convex path

Tags:

2018-09-18 About

A Theoretical Approach to Semantic Coding and Hashing | Simons Institute for the Theory of Computing (2016)

Tags:

2018-05-26 About

Comparing Sentence Similarity Methods

Tags:

2018-05-25 About

Named Entity Recognition and the Road to Deep Learning (2017)

Tags:

> the old
and the new-style NLP are not diametrically
opposed: just as it is possible (and useful!) to
incorporate neural-network features into a CRF,
CRFs have influenced some of the best deep
learning models for sequence labelling

This blog post go through the ways of doing NER, starting with CRF:

- When you develop a CRF,
a lot of time goes into finding feature functions (Does a word start with a capital? Is it uppercase? Is it a digit?...) 
- Pb: semantic similarity between words. 
- Gazetteers: lists with names of people, locations and organizations that are known in advance.
- feed Word Embeddings to a CRF: one way is to cluster a set
of word embeddings by distributional similarity, and
provide the CRF with the cluster IDs of a token and its
context words.
- Use word and character based embeddings
- LSTM not good enough -> biLSTM
- **biLSTM predict all labels independently of each other -> add a CRF layer** which outputs a matrix of transition scores between two states: dynamic programming can help find the optimal tag sequence for the sentence

2018-05-20 About

A Simple but Tough-to-Beat Baseline for Sentence Embeddings (2017)

Tags:

2018-05-10 About

[1801.06146] Universal Language Model Fine-tuning for Text Classification

Tags:

2018-01-19 About

Semantic Word Embeddings – Off the convex path

Tags:

2017-11-21 About

Semantic hashing (2008) - Ruslan Salakhutdinov, Geoffrey Hinton

Tags:

> We show how to learn a deep graphical model of the word-count vectors obtained from a
large set of documents. The values of the latent variables in the deepest layer are easy to
infer and give a much better representation of each document than Latent Semantic Analysis.
When the deepest layer is forced to use a small number of binary variables (e.g. 32),
the graphical model performs ‘‘semantic hashing”: Documents are mapped to memory
addresses in such a way that semantically similar documents are located at nearby
addresses. Documents similar to a query document can then be found by simply accessing
all the addresses that differ by only a few bits from the address of the query document. This
way of extending the efficiency of hash-coding to approximate matching is much faster
than locality sensitive hashing, which is the fastest current method. By using semantic
hashing to filter the documents given to TF-IDF, we achieve higher accuracy than applying
TF-IDF to the entire document set.

Indexing is implemented in the following manner: a document is mapped to a word-count vector and then this vector is passed through a [#Restricted Boltzmann Machine](/tag/restricted_boltzmann_machine) autoencoder and encoded to 32-bit address.

2017-11-07 About

Goldberg, Y. (2016). A Primer on Neural Network Models for Natural Language Processing. Journal of Artificial Intelligence Research

Tags:

2017-11-06 About

[1709.08568] The Consciousness Prior

Tags:

2017-09-29 About

An overview of word embeddings and their connection to distributional semantic models - AYLIEN (2016)

Tags:

2017-07-20 About

Quick review on Text Clustering and Text Similarity Approaches

Tags:

Author: Maali Mnasri (PhD @ CEA)

First transform text units to vectors? not always (eg. sentence similarity task using lexical word alignment). But vectors are efficient to process, and benefit from existing clustering algorithms such as k-means.

Sentence level or document level? Sentence clustering to summarise large documents.

Thematic clustering vs Semantic clustering: depends on the similarity measure.

Text similarity measures:

- Cosine similarity of tf-idf (suitable to produce thematic clusters)
- Knowledge-based Measures (wordNet) (quantify semantic relatedness of words),
- Word embedings

Examples, sample code:

- using wordnet with NLTK, and the formula to compute sentence similarities from word similarities.
- computing similarities between docs using gensim/word2vec

Which clustering algorithm?

- when we have an approximation of the clusters number, and when the similarity measure is not expensive in terms of computation time, clustering algo are suitable and fast. Sample code of k-means clustering using tf-idf vectors with scikit-learn
- Hierarchical clustering algorithms
    - don't need to give the number of clusters
    - but time consuming (calculate a similarity matrix for the sentences) 
- for voluminous data, use an incremental clustering algorithm: sentences are processed one at a time ; each new sentence is compared to each of the already formed clusters.

2017-05-18 About

Attention and Memory in Deep Learning and NLP – WildML

Tags:

cf. visual attention

In standard [#seq2seq](/tag/sequence_to_sequence_learning) NMT, the decoder is supposed to generate a translation solely based on the last hidden state of the encoder - which therefore must capture everything from the source sentence (it must be a sentence embedding). Not good. Hence the attention mechanism.

> we allow the decoder to “attend” to different parts of the source sentence at each step of the output generation. Importantly, we let the model learn what to attend to based on the input sentence and what it has produced so far

> each decoder output word now depends on a weighted combination of all the input states, not just the last state.

Possible to interpret what the model is doing by looking at the Attention weight matrix

Cost: We need to calculate an attention value for each combination of input and output word (-> attention is a bit of a misnomer: we look at everything in details before deciding what to focus on)

> attention mechanism is simply giving the network access to its internal memory, which is the hidden state of the encoder

> Unlike typical memory, the memory access mechanism here is soft, which means that the network retrieves a weighted combination of all memory locations, not a value from a single discrete location

2016-01-03 About

Deep Learning, NLP, and Representations - colah's blog

Tags:

2015-10-16 About

Introduction to Information Retrieval, Cambridge University Press (2008) Manning, Raghavan, and Schütze

Tags:

2012-04-01 About