widged/named-entity-recognition

1import streamlit as st2from transformers import pipeline3import spacy4from spacy import displacy5import plotly.express as px6import numpy as np7st.set_page_config(page_title="Named Entity Recognition")8st.title("Named Entity Recognition")9st.write("_This web application is intended for educational use, please do not upload any sensitive information._")10st.write("Identifying all geopolitical entities, organizations, people, locations, or dates in a body of text.")11 12@st.cache(allow_output_mutation=True, show_spinner=False)13def Loading_NLP():14    nlp = spacy.load('en_core_web_sm')15    return nlp16@st.cache(allow_output_mutation=True)17def entRecognizer(entDict, typeEnt):18    entList = [ent for ent in entDict if entDict[ent] == typeEnt]19    return entList20def plot_result(top_topics, scores):21    top_topics = np.array(top_topics)22    scores = np.array(scores)23    scores *= 10024    fig = px.bar(x=scores, y=top_topics, orientation='h',25                 labels={'x': 'Probability', 'y': 'Category'},26                 text=scores,27                 range_x=(0,115),28                 title='Top Predictions',29                 color=np.linspace(0,1,len(scores)),30                 color_continuous_scale="Bluered")31    fig.update(layout_coloraxis_showscale=False)32    fig.update_traces(texttemplate='%{text:0.1f}%', textposition='outside')33    st.plotly_chart(fig)34 35with st.spinner(text="Please wait for the models to load. This should take approximately 60 seconds."):36    nlp = Loading_NLP()37 38text = st.text_area('Enter Text Below:', height=300)39submit = st.button('Generate')40if submit:41    entities = []42    entityLabels = []43    doc = nlp(text)44    for ent in doc.ents:45        entities.append(ent.text)46        entityLabels.append(ent.label_)47    entDict = dict(zip(entities, entityLabels))48    entOrg = entRecognizer(entDict, "ORG")49    entPerson = entRecognizer(entDict, "PERSON")50    entDate = entRecognizer(entDict, "DATE")51    entGPE = entRecognizer(entDict, "GPE")52    entLoc = entRecognizer(entDict, "LOC")53    options = {"ents": ["ORG", "GPE", "PERSON", "LOC", "DATE"]}54    HTML_WRAPPER = """<div style="overflow-x: auto; border: 1px solid #e6e9ef; border-radius: 0.25rem; padding: 1rem; margin-bottom: 2.5rem">{}</div>"""55 56    st.subheader("List of Named Entities:")57    st.write("Geopolitical Entities (GPE): " + str(entGPE))58    st.write("People (PERSON): " + str(entPerson))59    st.write("Organizations (ORG): " + str(entOrg))60    st.write("Dates (DATE): " + str(entDate))61    st.write("Locations (LOC): " + str(entLoc))62    st.subheader("Original Text with Entities Highlighted")63    html = displacy.render(doc, style="ent", options=options)64    html = html.replace("\n", " ")65    st.write(HTML_WRAPPER.format(html), unsafe_allow_html=True)66