Ricerca semantica con vicini approssimativi e incorporamenti di testo

Visualizza su TensorFlow.org Esegui in Google Colab Visualizza su GitHub Scarica il taccuino Vedi il modello TF Hub

Questo tutorial illustra come generare incorporamenti da un modulo TensorFlow Hub (TF-Hub) dati di input e creare un indice approssimativo dei vicini più vicini (ANN) utilizzando gli incorporamenti estratti. L'indice può quindi essere utilizzato per la corrispondenza e il recupero delle somiglianze in tempo reale.

Quando si ha a che fare con un corpus di dati di grandi dimensioni, non è efficiente eseguire la corrispondenza esatta eseguendo la scansione dell'intero repository per trovare gli elementi più simili a una determinata query in tempo reale. Pertanto, utilizziamo un algoritmo di corrispondenza per similarità approssimativo che ci consente di sacrificare un po' di precisione nella ricerca delle corrispondenze esatte dei vicini più vicini con un significativo aumento della velocità.

In questo tutorial mostriamo un esempio di ricerca testuale in tempo reale su un corpus di titoli di notizie per trovare i titoli più simili a una query. A differenza della ricerca per parole chiave, questa cattura la somiglianza semantica codificata nell'incorporamento del testo.

I passaggi di questo tutorial sono:

  1. Scarica i dati di esempio.
  2. Genera incorporamenti per i dati utilizzando un modulo TF-Hub
  3. Costruisci un indice ANN per gli incorporamenti
  4. Utilizza l'indice per la corrispondenza delle somiglianze

Usiamo Apache Beam per generare gli incorporamenti dal modulo TF-Hub. Utilizziamo anche la libreria ANNOY di Spotify per creare l'indice approssimativo dei vicini più vicini.

Più modelli

Per i modelli che hanno la stessa architettura ma sono stati addestrati su un linguaggio diverso, fare riferimento a questa raccolta. Qui puoi trovare tutti gli incorporamenti di testo attualmente ospitati su tfhub.dev .

Impostare

Installa le librerie richieste.

pip install -q apache_beam
pip install -q 'scikit_learn~=0.23.0'  # For gaussian_random_matrix.
pip install -q annoy

Importa le librerie richieste

import os
import sys
import pickle
from collections import namedtuple
from datetime import datetime
import numpy as np
import apache_beam as beam
from apache_beam.transforms import util
import tensorflow as tf
import tensorflow_hub as hub
import annoy
from sklearn.random_projection import gaussian_random_matrix
print('TF version: {}'.format(tf.__version__))
print('TF-Hub version: {}'.format(hub.__version__))
print('Apache Beam version: {}'.format(beam.__version__))
TF version: 2.4.0
TF-Hub version: 0.11.0
Apache Beam version: 2.26.0

1. Scarica i dati di esempio

Il set di dati A Million News Headlines contiene titoli di notizie pubblicati in un periodo di 15 anni provenienti dalla rispettabile Australian Broadcasting Corp. (ABC). Questo set di dati di notizie contiene un resoconto storico riepilogativo di eventi degni di nota avvenuti nel mondo dall'inizio del 2003 alla fine del 2017, con un focus più granulare sull'Australia.

Formato : dati su due colonne separate da tabulazioni: 1) data di pubblicazione e 2) testo del titolo. A noi interessa solo il testo del titolo.

wget 'https://dataverse.harvard.edu/api/access/datafile/3450625?format=tab&gbrecs=true' -O raw.tsv
wc -l raw.tsv
head raw.tsv
--2021-01-07 12:50:08--  https://dataverse.harvard.edu/api/access/datafile/3450625?format=tab&gbrecs=true
Resolving dataverse.harvard.edu (dataverse.harvard.edu)... 206.191.184.198
Connecting to dataverse.harvard.edu (dataverse.harvard.edu)|206.191.184.198|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 57600231 (55M) [text/tab-separated-values]
Saving to: ‘raw.tsv’

raw.tsv             100%[===================>]  54.93M  14.7MB/s    in 4.4s    

2021-01-07 12:50:14 (12.4 MB/s) - ‘raw.tsv’ saved [57600231/57600231]

1103664 raw.tsv
publish_date    headline_text
20030219    "aba decides against community broadcasting licence"
20030219    "act fire witnesses must be aware of defamation"
20030219    "a g calls for infrastructure protection summit"
20030219    "air nz staff in aust strike for pay rise"
20030219    "air nz strike to affect australian travellers"
20030219    "ambitious olsson wins triple jump"
20030219    "antic delighted with record breaking barca"
20030219    "aussie qualifier stosur wastes four memphis match"
20030219    "aust addresses un security council over iraq"

Per semplicità, manteniamo solo il testo del titolo e rimuoviamo la data di pubblicazione

!rm -r corpus
!mkdir corpus

with open('corpus/text.txt', 'w') as out_file:
  with open('raw.tsv', 'r') as in_file:
    for line in in_file:
      headline = line.split('\t')[1].strip().strip('"')
      out_file.write(headline+"\n")
rm: cannot remove 'corpus': No such file or directory

tail corpus/text.txt
severe storms forecast for nye in south east queensland
snake catcher pleads for people not to kill reptiles
south australia prepares for party to welcome new year
strikers cool off the heat with big win in adelaide
stunning images from the sydney to hobart yacht
the ashes smiths warners near miss liven up boxing day test
timelapse: brisbanes new year fireworks
what 2017 meant to the kids of australia
what the papodopoulos meeting may mean for ausus
who is george papadopoulos the former trump campaign aide

2. Generare incorporamenti per i dati.

In questo tutorial utilizziamo il Neural Network Language Model (NNLM) per generare incorporamenti per i dati del titolo. Gli incorporamenti delle frasi possono quindi essere facilmente utilizzati per calcolare la somiglianza di significato a livello di frase. Eseguiamo il processo di generazione dell'incorporamento utilizzando Apache Beam.

Metodo di estrazione dell'incorporamento

embed_fn = None

def generate_embeddings(text, module_url, random_projection_matrix=None):
  # Beam will run this function in different processes that need to
  # import hub and load embed_fn (if not previously loaded)
  global embed_fn
  if embed_fn is None:
    embed_fn = hub.load(module_url)
  embedding = embed_fn(text).numpy()
  if random_projection_matrix is not None:
    embedding = embedding.dot(random_projection_matrix)
  return text, embedding

Converti nel metodo tf.Example

def to_tf_example(entries):
  examples = []

  text_list, embedding_list = entries
  for i in range(len(text_list)):
    text = text_list[i]
    embedding = embedding_list[i]

    features = {
        'text': tf.train.Feature(
            bytes_list=tf.train.BytesList(value=[text.encode('utf-8')])),
        'embedding': tf.train.Feature(
            float_list=tf.train.FloatList(value=embedding.tolist()))
    }

    example = tf.train.Example(
        features=tf.train.Features(
            feature=features)).SerializeToString(deterministic=True)

    examples.append(example)

  return examples

Conduttura del fascio

def run_hub2emb(args):
  '''Runs the embedding generation pipeline'''

  options = beam.options.pipeline_options.PipelineOptions(**args)
  args = namedtuple("options", args.keys())(*args.values())

  with beam.Pipeline(args.runner, options=options) as pipeline:
    (
        pipeline
        | 'Read sentences from files' >> beam.io.ReadFromText(
            file_pattern=args.data_dir)
        | 'Batch elements' >> util.BatchElements(
            min_batch_size=args.batch_size, max_batch_size=args.batch_size)
        | 'Generate embeddings' >> beam.Map(
            generate_embeddings, args.module_url, args.random_projection_matrix)
        | 'Encode to tf example' >> beam.FlatMap(to_tf_example)
        | 'Write to TFRecords files' >> beam.io.WriteToTFRecord(
            file_path_prefix='{}/emb'.format(args.output_dir),
            file_name_suffix='.tfrecords')
    )

Generazione della matrice del peso di proiezione casuale

La proiezione casuale è una tecnica semplice ma potente utilizzata per ridurre la dimensionalità di un insieme di punti che si trovano nello spazio euclideo. Per uno sfondo teorico, vedere il lemma di Johnson-Lindenstrauss .

Ridurre la dimensionalità degli incorporamenti con la proiezione casuale significa meno tempo necessario per costruire ed interrogare l'indice ANN.

In questo tutorial utilizziamo la proiezione casuale gaussiana dalla libreria Scikit-learn .

def generate_random_projection_weights(original_dim, projected_dim):
  random_projection_matrix = None
  random_projection_matrix = gaussian_random_matrix(
      n_components=projected_dim, n_features=original_dim).T
  print("A Gaussian random weight matrix was creates with shape of {}".format(random_projection_matrix.shape))
  print('Storing random projection matrix to disk...')
  with open('random_projection_matrix', 'wb') as handle:
    pickle.dump(random_projection_matrix, 
                handle, protocol=pickle.HIGHEST_PROTOCOL)

  return random_projection_matrix

Impostare i parametri

Se desideri creare un indice utilizzando lo spazio di incorporamento originale senza proiezione casuale, imposta il parametro projected_dim su None . Si noti che ciò rallenterà la fase di indicizzazione per gli incorporamenti ad alta dimensione.

Esegui la pipeline

import tempfile

output_dir = tempfile.mkdtemp()
original_dim = hub.load(module_url)(['']).shape[1]
random_projection_matrix = None

if projected_dim:
  random_projection_matrix = generate_random_projection_weights(
      original_dim, projected_dim)

args = {
    'job_name': 'hub2emb-{}'.format(datetime.utcnow().strftime('%y%m%d-%H%M%S')),
    'runner': 'DirectRunner',
    'batch_size': 1024,
    'data_dir': 'corpus/*.txt',
    'output_dir': output_dir,
    'module_url': module_url,
    'random_projection_matrix': random_projection_matrix,
}

print("Pipeline args are set.")
args
A Gaussian random weight matrix was creates with shape of (128, 64)
Storing random projection matrix to disk...
Pipeline args are set.

/tmpfs/src/tf_docs_env/lib/python3.6/site-packages/sklearn/utils/deprecation.py:86: FutureWarning: Function gaussian_random_matrix is deprecated; gaussian_random_matrix is deprecated in 0.22 and will be removed in version 0.24.
  warnings.warn(msg, category=FutureWarning)

{'job_name': 'hub2emb-210107-125029',
 'runner': 'DirectRunner',
 'batch_size': 1024,
 'data_dir': 'corpus/*.txt',
 'output_dir': '/tmp/tmp0g361gzp',
 'module_url': 'https://tfhub.dev/google/nnlm-en-dim128/2',
 'random_projection_matrix': array([[-0.1349755 , -0.12082699,  0.07092581, ..., -0.02680793,
         -0.0459312 , -0.20462361],
        [-0.06197901,  0.01832142,  0.21362496, ...,  0.06641898,
          0.14553738, -0.117217  ],
        [ 0.03452009,  0.14239163,  0.01371371, ...,  0.10422342,
          0.02966668, -0.07094185],
        ...,
        [ 0.03384223,  0.05102025,  0.01941788, ..., -0.07500625,
          0.09584965, -0.08593636],
        [ 0.11010087, -0.10597793,  0.06668758, ..., -0.0518654 ,
         -0.14681441,  0.08449293],
        [ 0.26909502, -0.0291555 ,  0.04305639, ..., -0.02295843,
          0.1164921 , -0.04828371]])}
print("Running pipeline...")
%time run_hub2emb(args)
print("Pipeline is done.")
WARNING:apache_beam.runners.interactive.interactive_environment:Dependencies required for Interactive Beam PCollection visualization are not available, please use: `pip install apache-beam[interactive]` to install necessary dependencies to enable all data visualization features.

Running pipeline...

Warning:tensorflow:5 out of the last 5 calls to <function recreate_function.<locals>.restored_function_body at 0x7efcac3599d8> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.

Warning:tensorflow:5 out of the last 5 calls to <function recreate_function.<locals>.restored_function_body at 0x7efcac3599d8> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.

Warning:tensorflow:6 out of the last 6 calls to <function recreate_function.<locals>.restored_function_body at 0x7efcac475598> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.

Warning:tensorflow:6 out of the last 6 calls to <function recreate_function.<locals>.restored_function_body at 0x7efcac475598> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.
WARNING:apache_beam.io.tfrecordio:Couldn't find python-snappy so the implementation of _TFRecordUtil._masked_crc32c is not as fast as it could be.

CPU times: user 9min 4s, sys: 10min 14s, total: 19min 19s
Wall time: 2min 30s
Pipeline is done.

ls {output_dir}
emb-00000-of-00001.tfrecords

Leggi alcuni degli incorporamenti generati...

embed_file = os.path.join(output_dir, 'emb-00000-of-00001.tfrecords')
sample = 5

# Create a description of the features.
feature_description = {
    'text': tf.io.FixedLenFeature([], tf.string),
    'embedding': tf.io.FixedLenFeature([projected_dim], tf.float32)
}

def _parse_example(example):
  # Parse the input `tf.Example` proto using the dictionary above.
  return tf.io.parse_single_example(example, feature_description)

dataset = tf.data.TFRecordDataset(embed_file)
for record in dataset.take(sample).map(_parse_example):
  print("{}: {}".format(record['text'].numpy().decode('utf-8'), record['embedding'].numpy()[:10]))
headline_text: [ 0.07743962 -0.10065071 -0.03604915  0.03902601  0.02538098 -0.01991337
 -0.11972483  0.03102058  0.16498186 -0.04299153]
aba decides against community broadcasting licence: [ 0.02420221 -0.07736929  0.05655728 -0.18739551  0.11344934  0.12652674
 -0.18189304  0.00422473  0.13149698  0.01910412]
act fire witnesses must be aware of defamation: [-0.17413895 -0.05418579  0.07769868  0.05096476  0.08622053  0.33112594
  0.04067763  0.00448784  0.15882017  0.33829722]
a g calls for infrastructure protection summit: [ 0.16939437 -0.18585566 -0.14201084 -0.21779229 -0.1374832   0.14933842
 -0.19583155  0.12921487  0.09811856  0.099967  ]
air nz staff in aust strike for pay rise: [ 0.0230642  -0.03269081  0.18271443  0.23761444 -0.01575144  0.06109515
 -0.01963143 -0.05211507  0.06050447 -0.20023327]

3. Costruire l'indice ANN per gli incorporamenti

ANNOY (Approximate Nearest Neighbors Oh Yeah) è una libreria C++ con collegamenti Python per cercare punti nello spazio vicini a un determinato punto di query. Crea inoltre grandi strutture di dati basate su file di sola lettura che vengono mappate in memoria. È costruito e utilizzato da Spotify per consigli musicali. Se sei interessato puoi giocare insieme ad altre alternative ad ANNOY come NGT , FAISS , ecc.

def build_index(embedding_files_pattern, index_filename, vector_length, 
    metric='angular', num_trees=100):
  '''Builds an ANNOY index'''

  annoy_index = annoy.AnnoyIndex(vector_length, metric=metric)
  # Mapping between the item and its identifier in the index
  mapping = {}

  embed_files = tf.io.gfile.glob(embedding_files_pattern)
  num_files = len(embed_files)
  print('Found {} embedding file(s).'.format(num_files))

  item_counter = 0
  for i, embed_file in enumerate(embed_files):
    print('Loading embeddings in file {} of {}...'.format(i+1, num_files))
    dataset = tf.data.TFRecordDataset(embed_file)
    for record in dataset.map(_parse_example):
      text = record['text'].numpy().decode("utf-8")
      embedding = record['embedding'].numpy()
      mapping[item_counter] = text
      annoy_index.add_item(item_counter, embedding)
      item_counter += 1
      if item_counter % 100000 == 0:
        print('{} items loaded to the index'.format(item_counter))

  print('A total of {} items added to the index'.format(item_counter))

  print('Building the index with {} trees...'.format(num_trees))
  annoy_index.build(n_trees=num_trees)
  print('Index is successfully built.')

  print('Saving index to disk...')
  annoy_index.save(index_filename)
  print('Index is saved to disk.')
  print("Index file size: {} GB".format(
    round(os.path.getsize(index_filename) / float(1024 ** 3), 2)))
  annoy_index.unload()

  print('Saving mapping to disk...')
  with open(index_filename + '.mapping', 'wb') as handle:
    pickle.dump(mapping, handle, protocol=pickle.HIGHEST_PROTOCOL)
  print('Mapping is saved to disk.')
  print("Mapping file size: {} MB".format(
    round(os.path.getsize(index_filename + '.mapping') / float(1024 ** 2), 2)))
embedding_files = "{}/emb-*.tfrecords".format(output_dir)
embedding_dimension = projected_dim
index_filename = "index"

!rm {index_filename}
!rm {index_filename}.mapping

%time build_index(embedding_files, index_filename, embedding_dimension)
rm: cannot remove 'index': No such file or directory
rm: cannot remove 'index.mapping': No such file or directory
Found 1 embedding file(s).
Loading embeddings in file 1 of 1...
100000 items loaded to the index
200000 items loaded to the index
300000 items loaded to the index
400000 items loaded to the index
500000 items loaded to the index
600000 items loaded to the index
700000 items loaded to the index
800000 items loaded to the index
900000 items loaded to the index
1000000 items loaded to the index
1100000 items loaded to the index
A total of 1103664 items added to the index
Building the index with 100 trees...
Index is successfully built.
Saving index to disk...
Index is saved to disk.
Index file size: 1.61 GB
Saving mapping to disk...
Mapping is saved to disk.
Mapping file size: 50.61 MB
CPU times: user 9min 54s, sys: 53.9 s, total: 10min 48s
Wall time: 5min 5s

ls
corpus         random_projection_matrix
index          raw.tsv
index.mapping  tf2_semantic_approximate_nearest_neighbors.ipynb

4. Utilizzare l'indice per la corrispondenza di somiglianza

Ora possiamo utilizzare l'indice ANN per trovare titoli di notizie semanticamente vicini a una query di input.

Caricare l'indice e i file di mappatura

index = annoy.AnnoyIndex(embedding_dimension)
index.load(index_filename, prefault=True)
print('Annoy index is loaded.')
with open(index_filename + '.mapping', 'rb') as handle:
  mapping = pickle.load(handle)
print('Mapping file is loaded.')
Annoy index is loaded.

/tmpfs/src/tf_docs_env/lib/python3.6/site-packages/ipykernel_launcher.py:1: FutureWarning: The default argument for metric will be removed in future version of Annoy. Please pass metric='angular' explicitly.
  """Entry point for launching an IPython kernel.

Mapping file is loaded.

Metodo di corrispondenza per somiglianza

def find_similar_items(embedding, num_matches=5):
  '''Finds similar items to a given embedding in the ANN index'''
  ids = index.get_nns_by_vector(
  embedding, num_matches, search_k=-1, include_distances=False)
  items = [mapping[i] for i in ids]
  return items

Estrai l'incorporamento da una determinata query

# Load the TF-Hub module
print("Loading the TF-Hub module...")
%time embed_fn = hub.load(module_url)
print("TF-Hub module is loaded.")

random_projection_matrix = None
if os.path.exists('random_projection_matrix'):
  print("Loading random projection matrix...")
  with open('random_projection_matrix', 'rb') as handle:
    random_projection_matrix = pickle.load(handle)
  print('random projection matrix is loaded.')

def extract_embeddings(query):
  '''Generates the embedding for the query'''
  query_embedding =  embed_fn([query])[0].numpy()
  if random_projection_matrix is not None:
    query_embedding = query_embedding.dot(random_projection_matrix)
  return query_embedding
Loading the TF-Hub module...
CPU times: user 757 ms, sys: 619 ms, total: 1.38 s
Wall time: 1.37 s
TF-Hub module is loaded.
Loading random projection matrix...
random projection matrix is loaded.

extract_embeddings("Hello Machine Learning!")[:10]
array([ 0.12164804,  0.0162079 , -0.15466002, -0.14580576,  0.03926325,
       -0.10124508, -0.1333948 ,  0.0515029 , -0.14688903, -0.09971556])

Inserisci una query per trovare gli articoli più simili

Generating embedding for the query...
CPU times: user 5.18 ms, sys: 596 µs, total: 5.77 ms
Wall time: 2.19 ms

Finding relevant items in the index...
CPU times: user 555 µs, sys: 327 µs, total: 882 µs
Wall time: 601 µs

Results:
=========
confronting global challenges
emerging nations to help struggling global economy
g7 warns of increasing global economic crisis
world struggling to cope with global terrorism
companies health to struggle amid global crisis
external risks biggest threat to economy
asian giants unite to tackle global crisis
g7 ministers warn of slowing global growth
experts to discuss global warming threat
scientists warn of growing natural disasters

Vuoi saperne di più?

Puoi saperne di più su TensorFlow su tensorflow.org e consultare la documentazione dell'API TF-Hub su tensorflow.org/hub . Trova i moduli TensorFlow Hub disponibili su tfhub.dev , inclusi altri moduli di incorporamento di testo e moduli vettoriali di funzionalità di immagine.

Dai un'occhiata anche al corso accelerato di Machine Learning , l'introduzione pratica e frenetica di Google all'apprendimento automatico.