app.py

import streamlit as st
from sentence_transformers.util import cos_sim
from sentence_transformers import models, SentenceTransformer,util
import pandas as pd
skills=['leading and motivating',
 'supervising people',
 'organising, planning and scheduling work and activities',
 'protecting and enforcing',
 'monitoring, inspecting and testing',
 'counselling',
 'using hand tools',
 'using digital tools for collaboration, content creation and problem solving',
 'teaching and training',
 'negotiating',
 'conducting studies, investigations and examinations',
 'handling and disposing of waste and hazardous materials',
 'developing objectives and strategies',
 'thinking creatively and innovatively',
 'working with others',
 'analysing and evaluating information and data',
 'installing interior or exterior infrastructure',
 'designing systems and products',
 'moving and lifting',
 'advising and consulting',
 'operating machinery for the manufacture of products',
 'handling animals',
 'setting up and protecting computer systems',
 'providing information and support to the public and clients',
 'writing and composing',
 'assembling and fabricating products',
 'operating mobile plant',
 'sorting and packaging goods and materials',
 'promoting, selling and purchasing',
 'providing health care or medical treatments',
 'documenting and recording information',
 'managing information',
 'accessing and analysing digital data',
 'preparing and serving food and drinks',
 'installing, maintaining and repairing mechanical equipment',
 'installing, maintaining and repairing electrical, electronic and precision equipment',
 'working with machinery and specialised equipment',
 'driving vehicles',
 'providing general personal care',
 'transforming and blending materials',
 'allocating and controlling resources',
 'liaising and networking',
 'performing and entertaining',
 'performing administrative activities',
 'monitoring developments in area of expertise',
 'tending plants and crops',
 'creating artistic, visual or instructive materials',
 'building and repairing structures',
 'processing information',
 'using precision instrumentation and equipment',
 'calculating and estimating',
 'obtaining information verbally',
 'presenting information',
 'making moulds, casts, models and patterns',
 'processing information, ideas and concepts',
 'making decisions',
 'dealing with problems',
 'solving problems',
 'measuring physical properties',
 'operating watercraft',
 'management skills',
 'positioning materials, tools or equipment',
 'washing and maintaining textiles and clothing',
 'operating machinery for the extraction and processing of raw materials',
 'working with computers',
 'information skills',
 'cleaning',
 'finishing interior or exterior of structures',
 'communication, collaboration and creativity',
 'constructing',
 'operating aircraft',
 'programming computer systems',
 'working in teams and collaboration',
 'using digital tools to control machinery',
 'assisting and caring',
 'using more than one language',
 'recruiting and hiring',
 'planning and organising',
 'handling and moving']


#mpnet_fine_tuned = SentenceTransformer('content/mpnet')
sbert_model = SentenceTransformer('h3lmi/fine_tuned_minilm12')
from sentence_transformers.util import cos_sim

def test(sentence,model,top_k=5):
    sims=[]
    skill_embeddings=model.encode(skills)
    query_vec=model.encode(sentence)
    hits = util.semantic_search(query_vec, skill_embeddings, top_k=top_k)[0]

    for hit in hits[0:top_k]:
        sims.append({'Skill': skills[hit['corpus_id']] ,'similarity':  hit['score']})

    
  
    return pd.DataFrame(sims)

st.title('Activity/Skill Semantic Matching')
search = st.text_input('Enter Activity/Skill')

if search:
    st.dataframe(test(search,sbert_model,5))