app.py

import streamlit as st
import spacy
from spacy_streamlit import visualize_ner
from support_functions import HealthseaPipe
import operator

# Header
with open("style.css") as f:
    st.markdown("<style>" + f.read() + "</style>", unsafe_allow_html=True)

# Intro
st.title("Welcome to Healthsea 🪐")

intro, jellyfish = st.columns(2)
jellyfish.markdown("\n")

intro.subheader("Create easier access to health✨")

jellyfish.image("data/img/Jellymation.gif")
intro.markdown(
    """Healthsea is an end-to-end spaCy v3 pipeline for analyzing user reviews to supplementary products and extracting their potential effects on health."""
)
intro.markdown(
    """The code for Healthsea is provided in this [github repository](https://github.com/thomashacker/healthsea). Visit our [blog post](https://explosion.ai/) or more about the Healthsea project.
    """
)

st.write(
    """This app visualizes the individual processing steps of the pipeline in which you can write custom reviews to get insights into the functionality of all the different components.
    You can visit the [Healthsea Demo app](https://huggingface.co/spaces/edichief/healthsea-demo) for exploring the Healthsea processing on productive data.
    """ 
)

st.markdown("""---""")

# Setup
healthsea_pipe = HealthseaPipe()

color_code = {
    "POSITIVE": ("#3C9E58", "#1B7735"),
    "NEGATIVE": ("#FF166A", "#C0094B"),
    "NEUTRAL": ("#7E7E7E", "#4E4747"),
    "ANAMNESIS": ("#E49A55", "#AD6B2D"),
}

example_reviews = [
    "This is great for joint pain.",
    "Product helped my joint pain but it also caused rashes.",
    "I'm diagnosed with gastritis. This product helped!",
    "This has made my insomnia even worse.",
    "It didn't help my joint pain.",
]

# Functions
def kpi(n, text):
    html = f"""
    <div class='kpi'>
        <h1>{n}</h1>
        <span>{text}</span>
    </div>
    """
    return html


def central_text(text):
    html = f"""<h2 class='central_text'>{text}</h2>"""
    return html


def format_clause(text, meta, pred):
    html = f"""
    <div>
        <div class="clause" style="background-color:{color_code[pred][0]} ; box-shadow: 0px 5px {color_code[pred][1]}; border-color:{color_code[pred][1]};">
            <div class="clause_text">{text}</div>
        </div>
        <div class="clause_meta">
            <div>{meta}</div>
        </div>
    </div>"""
    return html


def format_effect(text, pred):
    html = f"""
    <div>
        <div class="clause" style="background-color:{color_code[pred][0]} ; box-shadow: 0px 5px {color_code[pred][1]}; border-color:{color_code[pred][1]};">
            <div class="clause_text">{text}</div>
        </div>
    </div>"""
    return html

# Load model
load_state = st.info("Loading...")
try:
    load_state.info("Loading model...")
    if "model" not in st.session_state:
        nlp = spacy.load("en_healthsea")
        st.session_state["model"] = nlp
    load_state.success ("Loading complete!")

# Download model
except LookupError:
    import nltk
    import benepar
    load_state.info ("Downloading model...")
    benepar.download('benepar_en3')
    if "model" not in st.session_state:
        nlp = spacy.load("en_healthsea")
        st.session_state["model"] = nlp
    load_state.success ("Loading complete!")

except Exception as e:
    load_state.success ("Something went wrong!")
    st.error(e)

# Pipeline
st.markdown(central_text("⚙️ Pipeline"), unsafe_allow_html=True)

check = st.checkbox("Use predefined examples")

if not check:
    text = st.text_input(label="Write a review", value="This is great for joint pain!")
else:
    text = st.selectbox("Predefined example reviews", example_reviews)

nlp = st.session_state["model"]
doc = nlp(text)

# NER
visualize_ner(
    doc,
    labels=nlp.get_pipe("ner").labels,
    show_table=False,
    title="✨ Named Entity Recognition",
    colors={"CONDITION": "#FF4B76", "BENEFIT": "#629B68"},
)

st.info("""The NER identifies two labels: 'Condition' and 'Benefit'. 'Condition' entities are generally diseases, symptoms, or general health problems (e.g. joint pain), while 'Benefit' entities are positive desired health aspects (e.g. energy)""")

st.markdown("""---""")

# Segmentation, Blinding, Classification
st.markdown("## 🔮 Segmentation, Blinding, Classification")

clauses = healthsea_pipe.get_clauses(doc)
for doc_clause, clause in zip(clauses, doc._.clauses):
    classification = max(clause["cats"].items(), key=operator.itemgetter(1))[0]
    percentage = round(float(clause["cats"][classification]) * 100, 2)
    meta = f"{clause['ent_name']} ({classification} {percentage}%)"

    st.markdown(
        format_clause(doc_clause.text, meta, classification), unsafe_allow_html=True
    )
    st.markdown("\n")

st.info("""The text is segmented into clauses and classified by a Text Classification model. We additionally blind found entities to improve generalization and to inform the model about our current target entity.
The Text Classification predicts four exclusive classes that represent the health effect: 'Positive', 'Negative', 'Neutral', 'Anamnesis'.""")

st.info("""The 'Anamnesis' class is defined as the current state of health of a reviewer (e.g. 'I am diagnosed with joint pain'). It is used to link health aspects to health effects that are mentioned later in a review.""")

st.markdown("""---""")

# Aggregation
st.markdown("## 🔗 Aggregation")

for effect in doc._.health_effects:
    st.markdown(
        format_effect(
            f"{doc._.health_effects[effect]['effect']} effect on {effect}",
            doc._.health_effects[effect]["effect"],
        ),
        unsafe_allow_html=True,
    )
    st.markdown("\n")

st.info("""Multiple classification are aggregated into one final classification.""")

st.markdown("""---""")

# Indepth
st.markdown("## 🔧 Pipeline attributes")
clauses_col, effect_col = st.columns(2)

clauses_col.markdown("### doc._.clauses")
for clause in doc._.clauses:
    clauses_col.json(clause)
effect_col.markdown("### doc._.health_effects")
effect_col.json(doc._.health_effects)