Topic Discovery via Latent Space Clustering of Pretrained Language Model Representations

Last update: Dec 18, 2022

Overview

TopClus

The source code used for Topic Discovery via Latent Space Clustering of Pretrained Language Model Representations, published in WWW 2022.

Requirements

At least one GPU is required to run the code.

Before running, you need to first install the required packages by typing following commands (Using a virtual environment is recommended):

pip3 install -r requirements.txt

You need to also download the following resources in NLTK:

import nltk
nltk.download('stopwords')
nltk.download('averaged_perceptron_tagger')
nltk.download('universal_tagset')

Overview

TopClus is an unsupervised topic discovery method that jointly models words, documents and topics in a latent spherical space derived from pretrained language model representations.

Running Topic Discovery

The entry script is src/trainer.py and the meanings of the command line arguments will be displayed upon typing

python src/trainer.py -h

The topic discovery results will be written to results_${dataset}.

We provide two example scripts nyt.sh and yelp.sh for running topic discovery on the New York Times and the Yelp Review corpora used in the paper, respectively. You need to first extract the text files from the .tar.gz tarball files under datasets/nyt and datasets/yelp.

You could expect to obtain results like the following (the Topic IDs are random):

On New York Times:
Topic 20: months,weeks,days,decades,years,hours,decade,seconds,moments,minutes
Topic 28: weapons,missiles,missile,nuclear,grenades,explosions,explosives,launcher,bombs,bombing
Topic 30: healthcare,medical,medicine,physicians,patients,health,hospitals,bandages,medication,physician
Topic 41: economic,commercially,economy,business,industrial,industry,market,consumer,trade,commerce
Topic 46: senate,senator,congressional,legislators,legislatures,ministry,legislature,minister,ministerial,parliament
Topic 72: government,administration,governments,administrations,mayor,gubernatorial,mayoral,mayors,public,governor
Topic 77: aircraft,airline,airplane,airlines,voyage,airplanes,aviation,planes,spacecraft,flights
Topic 88: baseman,outfielder,baseball,innings,pitchers,softball,inning,basketball,shortstop,pitcher

On Yelp Review:
Topic 1: steamed,roasted,fried,shredded,seasoned,sliced,frozen,baked,canned,glazed
Topic 15: nice,cozy,elegant,polite,charming,relaxing,enjoyable,pleasant,helpful,luxurious
Topic 16: spicy,fresh,creamy,stale,bland,salty,fluffy,greasy,moist,cold
Topic 17: flavor,texture,flavors,taste,quality,smells,tastes,flavour,scent,ingredients
Topic 20: japanese,german,australian,moroccan,russian,greece,italian,greek,asian,
Topic 40: drinks,beers,beer,wine,beverages,alcohol,beverage,vodka,champagne,wines
Topic 55: horrible,terrible,shitty,awful,dreadful,worst,worse,disgusting,filthy,rotten
Topic 75: strawberry,berry,onion,peppers,tomato,onions,potatoes,vegetable,mustard,garlic

Running Document Clustering

The latent document embeddings will be saved to results_${dataset}/latent_doc_emb.pt which can be used as features to clustering algorithms (e.g., K-Means).

If you have ground truth document labels, you could obtain the document clustering evaluation results by passing the document label file and the saved latent document embedding file to the cluster_eval function in src/utils.py. For example:

from src.utils import TopClusUtils
utils = TopClusUtils()
utils.cluster_eval(label_path="datasets/nyt/label_topic.txt", emb_path="results_nyt/latent_doc_emb.pt")

Running on New Datasets

To execute the code on a new dataset, you need to

Create a directory named your_dataset under datasets.
Prepare a text corpus texts.txt (one document per line) under your_dataset as the target corpus for topic discovery.
Run src/trainer.py with appropriate command line arguments (the default values are usually good start points).

Citations

Please cite the following paper if you find the code helpful for your research.

@inproceedings{meng2022topic,
  title={Topic Discovery via Latent Space Clustering of Pretrained Language Model Representations},
  author={Meng, Yu and Zhang, Yunyi and Huang, Jiaxin and Zhang, Yu and Han, Jiawei},
  booktitle={The Web Conference},
  year={2022},
}

Topic Discovery via Latent Space Clustering of Pretrained Language Model Representations

Related tags

Overview

TopClus

Requirements

Overview

Running Topic Discovery

Running Document Clustering

Running on New Datasets

Citations

Owner

Yu Meng

PyTorch implementation of the TTC algorithm

An easier way to build neural search on the cloud

Diverse Object-Scene Compositions For Zero-Shot Action Recognition

PyTorch Implementation of Temporal Output Discrepancy for Active Learning, ICCV 2021

A demo of how to use JAX to create a simple gravity simulation

DCSAU-Net: A Deeper and More Compact Split-Attention U-Net for Medical Image Segmentation

Official code release for "GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis"

A toolset for creating Qualtrics-based IAT experiments

Code and data for "TURL: Table Understanding through Representation Learning"

Revisting Open World Object Detection

Code samples for my book "Neural Networks and Deep Learning"

Kindle is an easy model build package for PyTorch.

Code for sound field predictions in domains with impedance boundaries. Used for generating results from the paper

Bio-Computing Platform Featuring Large-Scale Representation Learning and Multi-Task Deep Learning “螺旋桨”生物计算工具集

Official code of "R2RNet: Low-light Image Enhancement via Real-low to Real-normal Network."

Collection of machine learning related notebooks to share.

Official Implementation of "LUNAR: Unifying Local Outlier Detection Methods via Graph Neural Networks"

Code for paper: Towards Tokenized Human Dynamics Representation

NeROIC: Neural Object Capture and Rendering from Online Image Collections

Audio-Visual Generalized Few-Shot Learning with Prototype-Based Co-Adaptation