Index

Search in book...
Toggle Font Controls
Create new playlist

Name your new playlist

Playlist description (optional)
Sign In

Email address

Password

Forgot Password?

or

Continue with Facebook

Continue with Google
Sign Up

Full Name

Email address

Confirm Email Address

Password

or

Continue with Facebook

Continue with Google

Previous Chapter

Chatbot Recirculating (Recurrent) Pipeline

List of Figures

Index

[SYMBOL][A][B][C][D][E][F][G][H][I][J][K][L][M][N][O][P][Q][R][S][T][U][V][W][X][Y][Z]

SYMBOL

@ operator 2-grams (bigrams) 2D thetas 2D vectors 3-grams (trigrams) 3D vectors, 2^nd 3x3 filter 4-grams (quadruplets) 4x4 matrix

A

ACK (acknowledgement) signal activation functions, 2^nd Adam additive smoothing AffNIST affordance AGI (artificial general intelligence), 2^nd AI (artificial intelligence), 2^nd AI Response Specification (AIRS) aichat bot framework aichat project AIML (Artificial Intelligence Markup Language), 2^nd pattern-matching chatbots with AIML 1.0 AIML 2.0 Python AIML interpreter <aiml> tags aiml_bot package AIRS (AI Response Specification) algorithms for scoring topics optimizing discretizing indexing search algorithms Amazon Alexa, 2^nd Amazon Echo Amazon Lex, 2^nd Amazon Web Services. See AWS. AMIs (Amazon Machine Images) Anaconda3 distribution analogies Analyze processing stage angular distance ANN (approximate nearest neighbors), 2^nd ANN (artificial neural network), 2^nd Annoy package Python Spotify AnnoyIndex object Apache Lucern + Solr API (application programmer interface) approximate grep approximate nearest neighbors. See ANN. apt package apt-get update command ARMA (auto-regressive moving average) model arrays artificial general intelligence. See AGI. artificial intelligence. See AI. artificial neural network. See ANN. attention mechanism augmenting data auto-regressive moving average model. See ARMA. autoencoders, 2^nd automata average option, pooling awards AWS (Amazon Web Services), 2^nd AWS (Amazon Web Services) GPU instances controlling costs creating AWS Billing Dashboard AWS Budget Console AWS Management Console Axes3D class axon

B

backpropagation, 2^nd, 3^rd, 4^th, 5^th, 6^th BallTree Basic Linear Algebra Subprograms library. See BLAS. Bastion host batch learning batch normalization batch_size, 2^nd Bayesian searches BeeSeek search engine bias avoiding backpropagation cost functions differentiable functions logical OR statements loss functions perceptrons limitations of training Pythonic neurons bias_weight feature bidirectional recurrent networks bigram scoring function bigrams, 2^nd Billing Dashboard, AWS binary_crossentropy function, 2^nd, 3^rd, 4^th BLAS (Basic Linear Algebra Subprograms) library, 2^nd bot-to-bot conversations BOW (bag-of-words), 2^nd, 3^rd, 4^th measuring overlap brittle pattern-matching Brown Corpus bucketing Budget Console, AWS built-in functions

C

Caffe framework candidate gate Cartesian distance case folding case normalization casual_tokenize function, 2^nd, 3^rd, 4^th categorical_crossentropy function, 2^nd CBOW (continuous bag-of-words), 2^nd CEC (constant error carousel) cells CFG (context-free grammars) chain rule character classes, 2^nd, 3^rd character prediction character range character sequence matches character-based model character-based translator chatbots, 2^nd approaches to assembling models for sequence generation building character dictionary building using FSM combining approaches connecting users conversing with design process emotional responses generating one-hot encoded training sets generating responses generative models grounding language skill community management conversational chatbots customer service hybrid approaches marketing chatbots modern approaches question answering dialog systems therapy virtual assistants non sequiturs pattern-matching approach network view of with AIML pipelines popular responses predicting sequences preparing corpus for training questions with predictable answers referring to search engines search-based context example of sequence-to-sequence, training Will chatbot framework ChatterBot checkpointing keyword argument Chloe, Aira.io class labels classifiers CNNs (convolutional neural networks), 2^nd architecture of filter composition frameworks for kernels narrow convolutions Dropout layers implementation in Keras models in pipelines pooling sigmoid activation function training models padding step size training word order CNTK environment co-occurrence matrix CodingBat collections module collections.Counter object collinear warning commas community management chatbots competitions compile method compiling models components attribute computational graphs Compute Unified Device Architecture. See CUDA. conda environment <condition> tag confusion matrix connectionist theory constant error carousel. See CEC. constant RAM algorithms gensim models graph computing context-free grammars. See CFG. contractions Conv1D layer convergence conversational chatbots, 2^nd convex error curve convolution tools, Keras convolutional neural networks. See CNNs. copy.copy() function CopyClip Core ML documentation, Apple CoreNLP library cosine distance, 2^nd, 3^rd, 4^th, 5^th cosine similarity, 2^nd, 3^rd cost functions Counter dictionary, 2^nd courses create_projection function CRM (customer relationship management system) cross-validation CUDA (Compute Unified Device Architecture) CuDNNLSTM curly braces, 2^nd customer relationship management system. See CRM. customer service chatbots

D

DAG (directed acyclic graph), 2^nd dark patterns dashes, 2^nd data augmenting labeled selecting data-driven programming DataCamp DataFrame, 2^nd datasets, 2^nd preparing for sequence-to-sequence training scaling shuffling dates, extracting dateutil.parser.parse DBpedia decode_sequence function decoder rings decoder_input_data decoder_outputs decoder_target_data decoding thought deep learning, 2^nd dendrites denominations denormalized data desired outcomes DetectorMorse deterministic finite automaton. See DFA. developer tools for Mac installing DFA (deterministic finite automaton), 2^nd dialog engines, 2^nd. See also chatbots. See also chatbots. dialog system DialogFlow, Google, 2^nd dict keys dict objects dictionaries, 2^nd differentiable functions dimension reduction dimensionality directed acyclic graph. See DAG. Dirichlet distribution discrete vectors discretizing distance functions distance metrics distances, 2^nd cosine distance Euclidean distance Manhattan distance distributed search engines Doc2vec algorithm Docker platform document vectors, 2^nd documenting similarity with Doc2vec documents, loading Donne Martin’s Coding Challenges dot products double quotes double-backslash dropout Dropout layers, 2^nd, 3^rd Dropout(percentage) Duck Duck Go search engine, 2^nd, 3^rd, 4^th Duplex system, Google

E

EarlyStopping edit distance, 2^nd eigenvalues Elastic Search ELIZA dialog engine, 2^nd embedding_dims value embeddings, 2^nd, 3^rd, 4^th encoder-decoder architecture autoencoders decoding thought end-of-sequence token entity names epochs parameter, 2^nd error carousel error derivative error matrix error surface errors, backpropagating Euclidean distance, 2^nd, 3^rd, 4^th Euclidean spaces evaluate method exclamation point Execute processing stage exp() function exploding gradient problem extracting dates GPS locations information, 2^nd numerical data relationships entity name normalization normalizing POS tagging segmentation sentence segmentation with regular expressions split word patterns

F

f-string template Facebook, 2^nd FALCONN (Fast Lookup of cosine and Other Nearest Neighbors) Fast Lookup of Approximate Nearest Neighbors. See FLANN. fastText library, 2^nd features, choosing feed-forward network, 2^nd, 3^rd feedback, steering with file openers filters composition of shapes of .findall() function finite state machine. See FSM. finite state transducer. See FST. fit method, 2^nd fit_generator method fit_transpose .fit() method, 2^nd, 3^rd flags FLANN (Fast Lookup of Approximate Nearest Neighbors) Flatten() function flip_sign function floating point values flush() method folds for loops forget gate forward pass fractal dimensions fractional distance frameworks for CNNs frequency based models frequency vectors from_records() function FSM (finite state machine), 2^nd, 3^rd, 4^th, 5^th alternatives to building chatbot using regular expressions FST (finite state transducer), 2^nd full text search fully connected network fuzzy matching, 2^nd fuzzywuzzy package

G

GANs (generative adversarial networks) gated recurrent unit. See GRU. gated recurrent units gem package generalization Generate processing stage generative adversarial networks. See GANs. generative models chatbot approaches transfer learning generators, 2^nd genetic algorithms gensim models gensim package gensim Vocab object gensim Word2vec model gensim.KeyedVectors model type gensim.KeyedVectors.most_similar() method gensim.LsiModel gensim.models.KeyedVector class gensim.models.LsiModel, 2^nd gensim.Word2vec modules geographic information system. See GIS. get() method Gigablast search engine, 2^nd GIS (geographic information system), 2^nd GitGUI glob patterns Global Vectors vs. Word2vec GlobalMaxPooling1D layer glossary acronyms terms GloVe (Global Vectors), 2^nd, 3^rd go_backwards keyword argument Google Adwords Auction case study Google Assistant, 2^nd Google Dialogflow Google Home Google News Google Ngram Viewer, 2^nd Google Now Google Scholar Google Translate pipeline Google’s Allo GPS locations, extracting GPUs (graphical processing units), 2^nd controlling costs of AWS GPU instances creating AWS GPU instances rental options renting vs. buying training NLP models on gradient descent grammar, 2^nd, 3^rd, 4^th graph computing graphemes graphical processing units. See GPUs. grep application grid searches grounding, 2^nd grouping parentheses groups GRU (gated recurrent unit), 2^nd

H

h5py package half pipes hand-coded algorithms hand-crafted weights handlebars hardcoded patterns hashes help() function hidden weights Hierarchical Navigable Small World. See HNSW. high performance computing. See HPC. high-dimensional continuous vectors high-dimensional data high-dimensional indexing high-dimensional space high-dimensional vectors hashes high-dimensional thinking 1D index 2D, 3D, 4D indexes vector space indexes HipChat, 2^nd HNSW (Hierarchical Navigable Small World), 2^nd Homebrew package manager homographs homonyms homophones horizontal scaling HPC (high performance computing), 2^nd, 3^rd hyperbolic tangent function hyperparameters, 2^nd, 3^rd, 4^th, 5^th, 6^th hyphens

I

i18n (internationalization) IDE (integrated development environment), 2^nd IDF (inverse document frequency) if statements imbalanced training set IMDB data IncrementalPCA model independent tokens index scan indexing 1D index 2D, 3D, 4D indexes benefits of high-dimensional open source full-text indexers vector space indexes with Annoy India Technical University. See ITU. indiscerniblity property inference stage information extraction, 2^nd information retrieval. See IR. init_sims method initial_state argument inner product input_characters input() function installing NLPIA Will chatbot framework integrated development environment. See IDE. internationalization. See i18n. inverse document frequency. See IDF. inverse proportionality inverted index, 2^nd IR (information retrieval), 2^nd, 3^rd ITU (India Technical University)

J

Jaccard distance junk mail jupyter project

K

K output nodes K-Dimensional Trees k-fold cross-validation k-means Kaggle competitions KD-Tree algorithm keras API Keras library, 2^nd, 3^rd CNNs with RNNs with sequence-to-sequence models in kernels, 2^nd key-value stores KeyedVectors object, 2^nd, 3^rd KeyedVectors.vocab dict keyword searches killer app, Word2vec knowledge base knowledge graph inference Kronuz/Xapiand

L

L1 regularization L2 regularization labeled data language modeling Laplace smoothing Lasagne library, 2^nd latent semantic analysis. See LSA. latent semantic indexing. See LSI. latent semantic information layer inputs, LSTM layers, pooling LDA (linear discriminant analysis), 2^nd, 3^rd, 4^th classifiers LDiA and LDiA (Latent Dirichlet allocation), 2^nd, 3^rd, 4^th as spam classifier for SMS messages history of LDA and LSA vs. ldia_topic_vectors matrix learning rate, 2^nd left singular vectors lemmatization, 2^nd Lemur Project Components Levenshtein distance, 2^nd Levenshtein’s algorithm lexers lexicons likeability, predicting limit keyword argument linear algebra, 2^nd, 3^rd linear discriminant analysis. See LDA. linearly separable data list comprehensions locality sensitive hashing. See LSH. location invariance log() function logical OR statements long short-term memory. See LSTM. loss functions, 2^nd, 3^rd low-dimensional continuous vectors lower() function LSA (latent semantic analysis), 2^nd, 3^rd, 4^th, 5^th enhancements to for spam classification LDiA vs. Word2vec vs. LSH (locality sensitive hashing), 2^nd, 3^rd, 4^th high-dimensional indexing high-dimensional vectors hashes high-dimensional thinking vector space indexes predicting likeability LSHash package LSI (latent semantic indexing), 2^nd LSTM (long short-term memory), 2^nd, 3^rd, 4^th backpropagation gated recurrent units generating novel text modeling language optimization of stacking layers training models tuning models unknown tokens using novel words with peephole connections Lua package

M

M neurons Mac OS (operating system) developer tools package manager tuneups machine learning avoiding bias batch normalization building models cross-validation data selection dropout imbalanced training sets augmenting data oversampling undersampling labeled data overfit performance metrics measuring classifier performance measuring regressor performance regularization underfit machine translation machine-generated text Mahalanobis distance Management Console, AWS Manhattan distance, 2^nd manipulative search engines mapping in Python marketing chatbots, 2^nd mathematical notation matplotlib package matrix orientation matrix product matrix, transpose of max pooling max_training_samples maxlen variable, 2^nd, 3^rd, 4^th mean squared error. See MSE. measuring bag of words performance of classifiers performance of regressors system capability Meld tool MIH (multi index hashing), 2^nd mini-batch training Minkowski distance MinMaxScaler Mitsuku chatbot MNIST dataset model parameters, 2^nd model training model.fit() function model.summary() function, 2^nd models assembling models for sequence generation batch normalization building compiling dropout in pipelines regularization sequence-to-sequence in Keras training, 2^nd, 3^rd on GPUs reducing memory footprint during training tuning momentum model morphemes, 2^nd, 3^rd most_common() function most_similar method MSE (mean squared error), 2^nd multi index hashing. See MIH. multi-resolution grid searches multidimensional semantic vectors multilayer perceptron

N

n-character grams n-grams, 2^nd, 3^rd, 4^th overview of stop words Naive Bayes algorithm, 2^nd named entities information extraction knowledge base Natural Language Forensics natural language generation. See NLG. natural language processing. See NLP. Natural Language Toolkit. See NLTK. natural language understanding. See NLU. negative argument negative class labels negative sampling NELL (Never Ending Language Learning), 2^nd, 3^rd nessvector, 2^nd net network neural networks bias backpropagation cost functions differentiable functions logical OR statements loss functions Pythonic neurons deep learning error surface in Python Keras mini-batch training nonconvex error curves normalization perceptrons stochastic gradient descent neuron, 2^nd Never Ending Language Learning. See NELL. new weights ngrams function NLG (natural language generation) NLP (natural language processing), 2^nd, 3^rd chatbot pipelines measuring system capability overview of building chatbots extracting numerical data parallelizing computations GPU rental options renting GPUs vs. buying Tensor processing units training NLP models on GPUs pipeline layers practical applications for programming languages vs. NLPIA automatic environment provisioning procedures installing nlpia package, 2^nd, 3^rd, 4^th, 5^th, 6^th, 7^th nlpia README file nlpia.data.loaders.get_data() method, 2^nd nlpia/data directory nlpiaenv environment NLTK (Natural Language Toolkit) nltk.casual_tokenizer nltk.tag package NLU (natural language understanding), 2^nd, 3^rd NMF (nonnegative matrix factorization), 2^nd NMSLIB (Non-Metric Space Library) non-ASCII characters nonconvex error curves nonlinear activation function nonlinear dimension reduction nonlinearly separable data nonnegative matrix factorization. See NMF. nonnegativity property nonspam SMS messages normalized dot product normalizing batches entity names relations vocabulary case folding lemmatization stemming use cases novel text np.matmul() function NULL characters num_bytes function num_encoder_tokens num_neurons layer num_neurons parameter num_rows function numerical data, extracting numerical perceptrons numpy arrays, 2^nd, 3^rd Nutch

O

O(N) algorithm objective function Okapi BM25 ranking function one-dimensional filter shape one-hot column vector one-hot encoded training sets one-hot encoded vectors one-hot row vector one-hot vectors one-way neural network Open Mind Common Sense Open Source Full Text Search Engine open source full-text indexers open source projects open source search engines OpenFST OpenStreetMap optimizing LSTM NLP algorithms discretizing indexing OR statement OR symbol, 2^nd ord values OrderedDict original weights output gate output nodes output words output_vocab_size output_vocabulary overfitting reducing training samples oversampling

P

package manager for Mac pad characters pad_sequences method padding Pandas DataFrames, 2^nd, 3^rd, 4^th parallelizing NLP computations GPU rental options renting GPUs vs. buying tensor processing units training NLP models on GPUs parentheses, 2^nd Parse processing stage parts of speech. See POS. pattern matching algorithms information extraction regular expressions pattern-matching chatbots, 2^nd, 3^rd network view of with AIML AIML 1.0 AIML 2.0 Python AIML interpreter pattern_response dictionary patterns of words PCA (principal component analysis), 2^nd, 3^rd, 4^th, 5^th, 6^th for SMS message semantic analysis LSA for spam classification LSA enhancements SVD enhancements NLP (natural language processing) on 3D vectors truncated SVD for SMS message semantic analysis PCA.fit() method pd.DataFrame printouts peephole connections PEP8 program perceptrons limitations of numerical training performance metrics measuring classifier performance measuring regressor performance periods phrase-terminating punctuation PhraseMatcher Pierson correlation pip (pip installs pip) pipelines chatbots convolutional layers of models in sequence-to-sequence, assembling assembling networks models in Keras preparing datasets for training sequence encoders thought decoders PiQASso Plotly wrapper plus sign Poisson distribution polarity polysemy pooling Porter stemmer POS (parts of speech), 2^nd, 3^rd, 4^th positive argument positive class labels PostgreSQL database, 2^nd PR (pull request) precision, 2^nd predicate predict method, 2^nd, 3^rd .predict_class() method predict_classes method .predict() method, 2^nd predicting bidirectional recurrent neural networks likeability sequences statefulness pretrained word vector principal component analysis. See PCA. probabilities_list function processing stages, chatbot product() function programming languages vs. NLP Project Gutenberg dataset project ideas projected distance propagation pull request. See PR. punctuation sequences pyfst interface Python programming language AIML interpreter dict objects mapping in mastering neural networks in neurons in OrderedDict strings and strings, types of style conventions templates in PyTorch framework, 2^nd, 3^rd, 4^th

Q

QDA (quadratic discriminant analysis), 2^nd quadruplets queries, semantic question answering (QA) systems, 2^nd, 3^rd quotes, 2^nd

R

random guessing random projection random searches <random> tag random values rare n-grams raw data raw strings RBMs (restricted Boltzmann machines) re package, 2^nd re.compile() function re.split function, 2^nd read–evaluate–print loop. See REPL. real-value indexes recall, 2^nd rectified linear unit. See ReLU. recurrent neural networks. See RNNs. Regex OR symbol regex package, 2^nd, 3^rd, 4^th RegexpTokenizer function regressors, measuring performance of regular expressions, 2^nd, 3^rd, 4^th character classes groups OR symbol overview of sentence segmentation with separating contractions separating words regularization relationships between words extracting entity name normalization normalizing POS tagging segmentation sentence segmentation with regular expressions split word patterns information extraction knowledge base relevance ranking ReLU (rectified linear unit), 2^nd, 3^rd renting GPUs buying vs. options for REPL (read–evaluate–print loop) response mappings restricted Boltzmann machines. See RBMs. retrieval-based chatbots. See also search-based chatbots. return_sequences keyword argument, 2^nd, 3^rd, 4^th return_state argument RMSE (root mean square error), 2^nd, 3^rd RNNs (recurrent neural networks), 2^nd, 3^rd backpropagation compiling models hyperparameters limitations of predicting bidirectional recurrent neural networks statefulness training models updating with Keras robot journalists RocketML pipelines root conda environment root mean square error. See RMSE. row vectors RSMProp rstr package rule-based algorithm

S

S matrix SAD (sum of absolute distance) saved models, loading scalability scalable vector graphics. See SVG. scalar product scaling datasets scanners scikit-learn package, 2^nd, 3^rd scikit-learn TruncatedSVD transformer scoring function scoring topics, algorithms for LDA LDiA search engines algorithms distributed less manipulative manipulative open source open source full-text indexers search-based chatbots context example of seed argument segment_sentences() function segmentation, 2^nd semantic analysis algorithms for scoring topics LDA LDiA distance LDA classifiers LDiA and LDiA as spam classifier for SMS messages history of LSA vs. LSA PCA for SMS message semantic analysis LSA for spam classification NLP (natural language processing) on 3D vectors truncated SVD for SMS message semantic analysis similarity steering with feedback SVD left singular vectors matrix orientation right singular vectors S matrix singular values truncating topics U matrix VT matrix TF-IDF vectors and lemmatization topic vectors, 2^nd semantic queries semantic searches, 2^nd, 3^rd semicolons sentence segmenting sentiment Naive Bayes model rule-based sentiment analyzers SentimentIntensityAnalyzer.lexicon separable data seq2seq (sequence-to-sequence networks) sequence decoders sequence encoders, 2^nd sequence-to-sequence applications for assembling networks assembling pipelines sequence encoders thought decoders building chatbots assembling models for sequence generation building character dictionary conversing with chatbots generating one-hot encoded training sets generating responses predicting sequences preparing corpus for training training sequence-to-sequence chatbots conversations encoder-decoder architecture autoencoders decoding thought models in Keras preparing datasets training training enhancements attention mechanism bucketing training networks sequences assembling models for sequence generation generating output sequences predicting Sequential class sequential minimal optimization. See SMO. Sequential() class Series object SGD (stochastic gradient descent), 2^nd, 3^rd .shape attribute shell commands sigmoid activation function fit method optimizers sigmoid function, 2^nd similarity, 2^nd simple_preprocess utility simpleNumericalFactChecker SimpleRNN layer, 2^nd, 3^rd, 4^th single-quotes singular value decomposition. See SVD. singular values Skflow skip-grams, 2^nd, 3^rd sklearn MinMaxScaler sklearn module, 2^nd, 3^rd, 4^th sklearn.decomposition.PCA sklearn.LatentDirichletAllocation sklearn.linear_model.Ridge regressor sklearn.manifold.TSNE sklearn.metrics.pairwise module sklearn.PCA model, 2^nd sklearn.Pipeline object sklearn.StandardScaler transform sklearn.TruncatedSVD Slack channels SMO (sequential minimal optimization) SMS messages LDiA for semantic analysis PCA for truncated SVD for Snappy Snowball stemmer softmax function learning vector representations overview of retrieving word vectors with linear algebra softmax layer, 2^nd softmax output value software patterns sorted() method spaces SpaCy package, 2^nd, 3^rd, 4^th, 5^th, 6^th, 7^th spacy.displacy spacy.matcher.Matcher spam classification, 2^nd spam dataset, SMS spam filters spam SMS messages sparse continuous vectors sparse matrices speech recognition speech to text. See STT. Sphinx Search split method split_turns function splitting, 2^nd spoofing square brackets, 2^nd squared Euclidean distance squashed vectors <srai> tag SSD (sum of squares distance) ssh credentials stacked layers stacking LSTM layers Stanford Core NLP library, 2^nd stanford-corenlp interface star character, 2^nd start token stateful keyword argument statefulness steering feature steering with feedback stemming, 2^nd algorithms for process of step function step size steps_per_epoch method stochastic gradient descent. See SGD. STOP token stop words stop_condition, 2^nd str class str.lower() function str.split() method, 2^nd, 3^rd stride string buffer string manipulation strings strip method STT (speech to text) style conventions Stylometry subject subsampling frequent tokens sum of absolute distance. See SAD. sum of squares distance. See SSD. .sum() method supervised algorithms SVD (singular value decomposition), 2^nd, 3^rd, 4^th enhancements to left singular vectors matrix orientation right singular vectors S matrix singular values truncated truncating topics U matrix VT matrix SVG (scalable vector graphics) SVM (support vector machine) symmetry property synonyms SyntaxNet package, 2^nd

T

T attribute t-Distributed Stochastic Neighbor Embedding (t-SNE), 2^nd tanh activation function target labels target (output) variable target unzipper target values target words target_seq, 2^nd target_text tdm term-document matrix teacher forcing method templates in Python tensor processing units. See TPUs. TensorBoard, 2^nd TensorFlow, 2^nd, 3^rd, 4^th, 5^th, 6^th term frequency. See TF. term frequency times inverse document frequency. See TF-IDF. term frequency vectors term-document matrices term-topic matrix terminals test set text characters text retrieval TF (term frequency) TF vectors TF-IDF (term frequency times inverse document frequency), 2^nd bag of words lemmatization and vectors topic modeling alternatives to Okapi BM25 relevance ranking tools for Zipf‘s Law vectorizing Zipf‘s Law tfidf vector TfidfVectorizer TFIDFVectorizer model, 2^nd Theano, 2^nd, 3^rd therapy chatbots thinking, high-dimensional 1D index 2D, 3D, 4D indexes thought decoders, 2^nd thought encoders threshold function, 2^nd time series data Time Series Matching time step tmp directory .todense() method token step token-by-token prediction tokenized phrases tokenizers dot products measuring bag-of-words overlap n-grams overview of stop words normalizing vocabulary case folding lemmatization stemming use cases regular expressions overview of separating contractions separating words tokens, 2^nd morphology of subsampling topic modeling alternatives to Okapi BM25 relevance ranking tools for Zipf‘s Law topic vectors, 2^nd topic weights topics, 2^nd topn argument Torch package TPUs (tensor processing units) train_test_split() method train/ directory trained model, 2^nd training CNNs document vectors domain-specific Word2vec models enhancements for attention mechanism bucketing models, 2^nd, 3^rd on GPUs reducing memory footprint during training perceptrons preparing corpus for preparing datsets for sequence-to-sequence training sequence-to-sequence chatbots sequence-to-sequence networks training data, loading training methods training samples training sets, 2^nd imbalanced augmenting data oversampling undersampling training_set_generator function transfer learning .transform() method transpose of matrix Treebank tokenizer Treebank Word Tokenizer, 2^nd, 3^rd triangle inequality property trigger words trigrams triple-quoted raw strings troll message filtering truncated characters truncated data truncated singular value decomposition TruncatedSVD model, 2^nd truncating topics in SVD TurboTax, Intuit Turing test tutorials two-dimensional filter shape

U

U matrix Ubuntu Dialog Corpus, 2^nd Ubuntu package manager UI (user interface) underfitting, 2^nd undersampling underscore character, 2^nd Unicode characters unique tokens, 2^nd units, 2^nd UNK (unknown) token unnatural words unrolled net, 2^nd, 3^rd, 4^th update gate updating RNNs user interface. See UI. UX (user experience)

V

VADER algorithm vaderSentiment package validating dates validation set vanishing gradient problem variance, maximizing variational autoencoder vector difference vector dimensions vector representations vector space model. See VSM. vector spaces, 2^nd indexes splitting vector-oriented reasoning vectorizing vectors, 2^nd 3D vectors distances cosine distance Euclidean distance Manhattan distance high-dimensional hashes high-dimensional thinking vector space indexes left singular right singular virtual assistant chatbots virtual assistants virtual private cloud. See VPC. VirtualBox application visualize_embeddings function vocabulary, normalizing case folding lemmatization stemming use cases VPC (virtual private cloud) VSM (vector space model), 2^nd, 3^rd VT matrix

W

Watson Semantic Web Search WebSphinx search engine weights, 2^nd, 3^rd, 4^th, 5^th, 6^th whitespace, 2^nd Whoosh, Python Wikia Search Wikidata Will chatbot framework installing untrained word embeddings, 2^nd word frequencies word order, 2^nd word prediction word proximity word tokenization challenges of sentiment Naive Bayes model rule-based sentiment analyzers tokenizers dot products measuring bag-of-words overlap n-grams normalizing vocabulary regular expressions word vectors analogies computing Word2vec representations CBOW approach frequent bigrams negative sampling skip-gram approach skip-gram vs. CBOW softmax function subsampling frequent tokens documenting similarity with Doc2vec embedding fastText generating word vector representations preprocessing steps training domain-specific Word2vec models gensim.Word2vec modules retrieving with linear algebra semantic queries unnatural words vector-oriented reasoning visualizing word relationships word-topic vectors Word2vec, 2^nd, 3^rd, 4^th, 5^th computing representations CBOW approach frequent bigrams negative sampling skip-gram approach skip-gram vs. CBOW softmax function subsampling frequent tokens gensim.Word2vec modules GloVe vs. LSA vs. training domain-specific models WordNetLemmatizer words order of separating .words() method Wysa chatbot

X

Y

Yacy search engine YAGO Yandex Your Dictionary YourDOST

Z

zero determinant zero vector, 2^nd Zettair indexer zeugma Zipf‘s Law, 2^nd

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.