SIFRank serving

see original paper SIFRank: A New Baseline for Unsupervised Keyphrase Extraction Based on Pre-trained Language Model

Getting Started

docker-compose parameters

Flask api running on port 5000 will be mapped to outer 5001 port. (it uses docker-compose version 2.3 which supports runtime: nvidia to easily use GPU environment inside container)

version: '2.3'
services:
  sifrank_serving:
    container_name: sifrank_serving
    build: .
    ports:
      - "5001:5000"
    runtime: nvidia
    environment:
      - NVIDIA_VISIBLE_DEVICES=all

Running Docker Container

Probably the easiest way to get started is by using the provided Docker image. From the project's root directory, the image can be built and running like so:

$ docker-compose up --build -d

This can take a several minutes to finish. All model files and everything will be downloaded automatically.

Flask API (inside container)

This return for each text a tuple containing three lists:

1. The top N candidates (string) i.e keyphrases
2. For each keyphrase the associated relevance score

import nltk
import stanza
from flask import Flask, jsonify, request
from waitress import serve

from embeddings import sent_emb_sif, word_emb_elmo
from model.method import SIFRank, SIFRank_plus

app = Flask(__name__)


@app.route('/sifrank', methods=['POST'])
def sifrank():
    data = request.json
    query = data.get("text", "")
    top_n = int(data.get("n", 15))

    keywords = SIFRank(query, SIF, en_model, N=top_n, elmo_layers_weight=elmo_layers_weight)

    return jsonify(keywords)


@app.route('/sifrankplus', methods=['POST'])
def sifrankplus():
    data = request.json
    query = data.get("text", "")
    top_n = int(data.get("n", 15))

    keywords = SIFRank_plus(query, SIF, en_model, N=top_n, elmo_layers_weight=elmo_layers_weight)

    return jsonify(keywords)


if __name__ == '__main__':
    data_dir = "data/"
    options_file = f"{data_dir}elmo_2x4096_512_2048cnn_2xhighway_options.json"
    weight_file = f"{data_dir}elmo_2x4096_512_2048cnn_2xhighway_weights.hdf5"
    cuda_device = 0

    porter = nltk.PorterStemmer()
    ELMO = word_emb_elmo.WordEmbeddings(options_file=options_file, weight_file=weight_file, cuda_device=cuda_device)
    SIF = sent_emb_sif.SentEmbeddings(word_embeddor=ELMO, data_dir=data_dir, lamda=1.0)
    en_model = stanza.Pipeline(lang='en', processors={}, use_gpu=True)
    elmo_layers_weight = [0.0, 1.0, 0.0]

    serve(app, host="0.0.0.0", port=5000)

Curl example to call API

You can simply query from terminal

curl --location --request POST 'http://0.0.0.0:5001/sifrank' \            
    --header 'Content-Type: application/json' \
    --data-raw '{"text": "A New Baseline for Unsupervised Keyphrase Extraction Based on Pre-trained Language Model.", "n":10}'
    
curl --location --request POST 'http://0.0.0.0:5001/sifrankplus' \            
    --header 'Content-Type: application/json' \
    --data-raw '{"text": "A New Baseline for Unsupervised Keyphrase Extraction Based on Pre-trained Language Model.", "n":10}'

Response

[["pre-trained language model","new baseline","keyphrase extraction"],[0.91609799642446,0.8865735384713,0.8844055415033372]]
[["new baseline","pre-trained language model","keyphrase extraction"],[0.9202464080510155,0.88569978675073,0.8801616342917246]]

POST request from python

import json
import requests

sifrank_url = f"http://0.0.0.0:5001/sifrank"
sifrank_plus_url = f"http://0.0.0.0:5001/sifrankplus"

data = {"text": "A New Baseline for Unsupervised Keyphrase Extraction Based on Pre-trained Language Model.", "lang":"en", "n":10}
sifrank_result = requests.post(sifrank_url, json=data)
sifrank_plus_result = requests.post(sifrank_plus_url, json=data)

sifrank_content = json.loads(sifrank_result.content)
sifrank_plus_content = json.loads(sifrank_plus_result.content)

sifrank_keywords = sifrank_content[0]
sifrank_relevance_scores = sifrank_content[1]

sifrank_plus_keywords = sifrank_plus_content[0]
sifrank_plus_relevance_scores = sifrank_plus_content[1]

Local Run

requirements

allennlp==0.9.0
stanza==1.0.0
waitress==1.4.3

Download Models

• ELMo elmo_2x4096_512_2048cnn_2xhighway_options.json (already downloaded in data/) and elmo_2x4096_512_2048cnn_2xhighway_weights.hdf5 from here , and save it to the data/ directory

Sample usage

import nltk
from embeddings import sent_emb_sif, word_emb_elmo
from model.method import SIFRank, SIFRank_plus
import stanza

# download from https://allennlp.org/elmo
options_file = "data/elmo_2x4096_512_2048cnn_2xhighway_options.json"
weight_file = "data/elmo_2x4096_512_2048cnn_2xhighway_weights.hdf5"

porter = nltk.PorterStemmer()
ELMO = word_emb_elmo.WordEmbeddings(options_file, weight_file, cuda_device=0)
SIF = sent_emb_sif.SentEmbeddings(ELMO, lamda=1.0)
en_model = stanza.Pipeline(lang='en', processors={}, use_gpu=True)
elmo_layers_weight = [0.0, 1.0, 0.0]

text = "Discrete output feedback sliding mode control of second order systems - a moving switching line approach The sliding mode control systems (SMCS) for which the switching variable is designed independent of the initial conditions are known to be sensitive to parameter variations and extraneous disturbances during the reaching phase. For second order systems this drawback is eliminated by using the moving switching line technique where the switching line is initially designed to pass the initial conditions and is subsequently moved towards a predetermined switching line. In this paper, we make use of the above idea of moving switching line together with the reaching law approach to design a discrete output feedback sliding mode control. The main contributions of this work are such that we do not require to use system states as it makes use of only the output samples for designing the controller. and by using the moving switching line a low sensitivity system is obtained through shortening the reaching phase. Simulation results show that the fast output sampling feedback guarantees sliding motion similar to that obtained using state feedback"
keyphrases = SIFRank(text, SIF, en_model, N=15, elmo_layers_weight=elmo_layers_weight)
keyphrases_ = SIFRank_plus(text, SIF, en_model, N=15, elmo_layers_weight=elmo_layers_weight)
print(keyphrases)
print(keyphrases_)

SIFRank evaluation results from original repository

F1 score when the number of keyphrases extracted N is set to 5.

Models	Inspec	SemEval2017	DUC2001
TFIDF	11.28	12.70	9.21
YAKE	15.73	11.84	10.61
TextRank	24.39	16.43	13.94
SingleRank	24.69	18.23	21.56
TopicRank	22.76	17.10	20.37
PositionRank	25.19	18.23	24.95
Multipartite	23.05	17.39	21.86
RVA	21.91	19.59	20.32
EmbedRank d2v	27.20	20.21	21.74
SIFRank	29.11	22.59	24.27
SIFRank+	28.49	21.53	30.88

Cite

If you use this code, please cite this paper

@article{DBLP:journals/access/SunQZWZ20,
  author    = {Yi Sun and
               Hangping Qiu and
               Yu Zheng and
               Zhongwei Wang and
               Chaoran Zhang},
  title     = {SIFRank: {A} New Baseline for Unsupervised Keyphrase Extraction Based
               on Pre-Trained Language Model},
  journal   = {{IEEE} Access},
  volume    = {8},
  pages     = {10896--10906},
  year      = {2020},
  url       = {https://doi.org/10.1109/ACCESS.2020.2965087},
  doi       = {10.1109/ACCESS.2020.2965087},
  timestamp = {Fri, 07 Feb 2020 12:04:22 +0100},
  biburl    = {https://dblp.org/rec/journals/access/SunQZWZ20.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}