app.py

import nltk
import re
import nltkmodule

from nltk.tokenize import word_tokenize
from sentence_transformers import SentenceTransformer
import pandas as pd
import numpy as np
from pandas import ExcelWriter
from torch.utils.data import DataLoader
import math
from sentence_transformers import models, losses
from sentence_transformers import SentencesDataset, LoggingHandler, SentenceTransformer
from sentence_transformers.evaluation import EmbeddingSimilarityEvaluator
from sentence_transformers.readers import *
import logging
import glob
from datetime import datetime
import sys
from nltk.corpus import stopwords
stop_words = stopwords.words('english')
from sklearn.metrics.pairwise import cosine_similarity
import scipy.spatial
import networkx as nx
from nltk.tokenize import sent_tokenize
import scispacy
import spacy
import en_core_sci_lg
from spacy import displacy
from scispacy.abbreviation import AbbreviationDetector
from scispacy.umls_linking import UmlsEntityLinker
from transformers import AutoTokenizer, AutoModel
import statistics
import string
from nltk.stem.wordnet import WordNetLemmatizer
import gradio as gr

nlp = en_core_sci_lg.load()
sp = en_core_sci_lg.load()
all_stopwords = sp.Defaults.stop_words


def remove_stopwords(sen):
    sen_new = " ".join([i for i in sen if i not in stop_words])
    return sen_new

def keyphrase_generator(article, model_1, model_2, max_num_keywords):
  element=[]
  document=[]
  text=[]
  model_1 = SentenceTransformer(model_1)
  model_2 = SentenceTransformer(model_2)
  corpus=sent_tokenize(article)
  clean_sentences_new = pd.Series(corpus).str.replace("[^a-zA-Z]", " ").tolist()
  corpus_embeddings = model_1.encode(clean_sentences_new)
  sim_mat = np.zeros([len(clean_sentences_new), len(clean_sentences_new)])
  for i in range(len(clean_sentences_new)):
    for j in range(len(clean_sentences_new)):
      if i != j:
        sim_mat[i][j] = cosine_similarity(corpus_embeddings[i].reshape(1,768), corpus_embeddings[j].reshape(1,768))[0,0]
  nx_graph = nx.from_numpy_array(sim_mat)
  scores = nx.pagerank(nx_graph)
  ranked_sentences = sorted(((scores[i],s) for i,s in enumerate(corpus)), reverse=True)
  for elem in ranked_sentences:
    element.append(elem[1])
  a=int((10*len(element))/100.0)
  if(a<5):
    total=5
  else:
    total=int(a)
  for i in range(total):
    document.append(element[i])
  doc=" ".join(document)
  for i in document:
    doc_1=nlp(i)
    text.append([X.text for X in doc_1.ents])
  entity_list = [item for sublist in text for item in sublist]
  entity_list = [word for word in entity_list if not word in all_stopwords]
  entity_list=list(dict.fromkeys(entity_list))
  doc_embedding = model_2.encode([doc])
  candidates=entity_list
  candidate_embeddings = model_2.encode(candidates)
  distances = cosine_similarity(doc_embedding, candidate_embeddings)
  top_n = max_num_keywords
  keyword_list = [candidates[index] for index in distances.argsort()[0][-top_n:]]
  keywords = '\n'.join(keyword_list) 
  return keywords


gr.Interface(keyphrase_generator, 
             inputs=[gr.inputs.Textbox(lines=10, placeholder="Copy article text here",default="", label="article text"),gr.inputs.Textbox(lines=1, placeholder="SBERT model",default="all-mpnet-base-v2", label="Model for TextRank (e.g. all-mpnet-base-v2)"),gr.inputs.Textbox(lines=1, placeholder="SBERT model",default="all-distilroberta-v1",label="Model for keyphrases (e.g. all-distilroberta-v1)"),gr.inputs.Slider(minimum=5, maximum=30, step=1, default=10, label="Max Keywords")], 
             outputs="text", theme=None, title="Scientifc Article Keyphrase Generator", article="Generates the keyphrases from an article which best describes the article."
             "\t The work is part of the paper ""." 
             "\t It uses the TextRank algorithm to first find the top sentences and then extracts the keyphrases from those sentences."
             "\t The list of SBERT models required in the textboxes can be found in https://www.sbert.net/docs/pretrained_models.html."
             "\t The default model names are provided which can be changed from the list of pretrained models. "
             "\t The value of output keyphrases can be changed. The default value is 10, minimum is 5 and a maximum value of 30.").launch(share=True)