Upload folder using huggingface_hub

app.py

@@ -11,10 +11,10 @@ import pandas as pd
 from src.extractor import EntityExtractor
 from src.graph_builder import KnowledgeGraph, ENTITY_COLORS
 from src.visualizer import (
- create_pyvis_graph,
- graph_stats_chart,
- centrality_chart,
- community_chart,
+    create_pyvis_graph,
+    graph_stats_chart,
+    centrality_chart,
+    community_chart,
 )
 from src.sample_texts import SAMPLE_TEXTS
 
@@ -23,10 +23,10 @@ from src.sample_texts import SAMPLE_TEXTS
 # ======================================================================
 
 st.set_page_config(
- page_title="GraphMind | Knowledge Graph",
- page_icon="G",
- layout="wide",
- initial_sidebar_state="expanded",
+    page_title="GraphMind | Knowledge Graph",
+    page_icon="G",
+    layout="wide",
+    initial_sidebar_state="expanded",
 )
 
 # ======================================================================
@@ -34,111 +34,111 @@ st.set_page_config(
 # ======================================================================
 
 st.markdown(
- """
- <style>
- /* ---- Global ---- */
- .stApp {
- background-color: #0a0a0a;
- color: #e0e0e0;
- }
-
- /* ---- Sidebar ---- */
- section[data-testid="stSidebar"] {
- background-color: #111111;
- border-right: 1px solid #1e1e1e;
- }
-
- /* ---- Headers ---- */
- h1, h2, h3, h4 {
- color: #ffffff !important;
- }
-
- /* ---- Metric cards ---- */
- div[data-testid="stMetric"] {
- background: linear-gradient(135deg, #111111 0%, #1a1a2e 100%);
- border: 1px solid #1e1e1e;
- border-radius: 12px;
- padding: 16px 20px;
- }
- div[data-testid="stMetric"] label {
- color: #888888 !important;
- }
- div[data-testid="stMetric"] div[data-testid="stMetricValue"] {
- color: #00ff88 !important;
- font-weight: 700;
- }
-
- /* ---- Buttons ---- */
- .stButton > button {
- background: linear-gradient(135deg, #00ff88 0%, #00d4ff 100%);
- color: #0a0a0a;
- border: none;
- border-radius: 8px;
- font-weight: 700;
- padding: 0.5rem 1.5rem;
- transition: all 0.3s ease;
- }
- .stButton > button:hover {
- transform: translateY(-2px);
- box-shadow: 0 4px 20px rgba(0,255,136,0.3);
- }
-
- /* ---- Tabs ---- */
- .stTabs [data-baseweb="tab-list"] {
- gap: 8px;
- }
- .stTabs [data-baseweb="tab"] {
- background-color: #1a1a1a;
- border-radius: 8px 8px 0 0;
- color: #888888;
- padding: 8px 20px;
- }
- .stTabs [aria-selected="true"] {
- background-color: #1e1e2e;
- color: #00ff88 !important;
- }
-
- /* ---- DataFrame ---- */
- .stDataFrame {
- border: 1px solid #1e1e1e;
- border-radius: 8px;
- }
-
- /* ---- Expanders ---- */
- .streamlit-expanderHeader {
- background-color: #111111;
- border-radius: 8px;
- }
-
- /* ---- Success / info banners ---- */
- .stAlert {
- background-color: #111111;
- border: 1px solid #1e1e1e;
- border-radius: 8px;
- }
-
- /* ---- Accent text helpers ---- */
- .accent-green { color: #00ff88; font-weight: 700; }
- .accent-blue { color: #00d4ff; font-weight: 700; }
-
- /* ---- Legend colour pills ---- */
- .legend-pill {
- display: inline-block;
- padding: 3px 12px;
- border-radius: 20px;
- margin: 2px 4px;
- font-size: 0.82rem;
- font-weight: 600;
- color: #0a0a0a;
- }
-
- /* ---- Divider ---- */
- hr {
- border-color: #1e1e1e;
- }
- </style>
- """,
- unsafe_allow_html=True,
+    """
+    <style>
+    /* ---- Global ---- */
+    .stApp {
+        background-color: #0a0a0a;
+        color: #e0e0e0;
+    }
+
+    /* ---- Sidebar ---- */
+    section[data-testid="stSidebar"] {
+        background-color: #111111;
+        border-right: 1px solid #1e1e1e;
+    }
+
+    /* ---- Headers ---- */
+    h1, h2, h3, h4 {
+        color: #ffffff !important;
+    }
+
+    /* ---- Metric cards ---- */
+    div[data-testid="stMetric"] {
+        background: linear-gradient(135deg, #111111 0%, #1a1a2e 100%);
+        border: 1px solid #1e1e1e;
+        border-radius: 12px;
+        padding: 16px 20px;
+    }
+    div[data-testid="stMetric"] label {
+        color: #888888 !important;
+    }
+    div[data-testid="stMetric"] div[data-testid="stMetricValue"] {
+        color: #00ff88 !important;
+        font-weight: 700;
+    }
+
+    /* ---- Buttons ---- */
+    .stButton > button {
+        background: linear-gradient(135deg, #00ff88 0%, #00d4ff 100%);
+        color: #0a0a0a;
+        border: none;
+        border-radius: 8px;
+        font-weight: 700;
+        padding: 0.5rem 1.5rem;
+        transition: all 0.3s ease;
+    }
+    .stButton > button:hover {
+        transform: translateY(-2px);
+        box-shadow: 0 4px 20px rgba(0,255,136,0.3);
+    }
+
+    /* ---- Tabs ---- */
+    .stTabs [data-baseweb="tab-list"] {
+        gap: 8px;
+    }
+    .stTabs [data-baseweb="tab"] {
+        background-color: #1a1a1a;
+        border-radius: 8px 8px 0 0;
+        color: #888888;
+        padding: 8px 20px;
+    }
+    .stTabs [aria-selected="true"] {
+        background-color: #1e1e2e;
+        color: #00ff88 !important;
+    }
+
+    /* ---- DataFrame ---- */
+    .stDataFrame {
+        border: 1px solid #1e1e1e;
+        border-radius: 8px;
+    }
+
+    /* ---- Expanders ---- */
+    .streamlit-expanderHeader {
+        background-color: #111111;
+        border-radius: 8px;
+    }
+
+    /* ---- Success / info banners ---- */
+    .stAlert {
+        background-color: #111111;
+        border: 1px solid #1e1e1e;
+        border-radius: 8px;
+    }
+
+    /* ---- Accent text helpers ---- */
+    .accent-green { color: #00ff88; font-weight: 700; }
+    .accent-blue  { color: #00d4ff; font-weight: 700; }
+
+    /* ---- Legend colour pills ---- */
+    .legend-pill {
+        display: inline-block;
+        padding: 3px 12px;
+        border-radius: 20px;
+        margin: 2px 4px;
+        font-size: 0.82rem;
+        font-weight: 600;
+        color: #0a0a0a;
+    }
+
+    /* ---- Divider ---- */
+    hr {
+        border-color: #1e1e1e;
+    }
+    </style>
+    """,
+    unsafe_allow_html=True,
 )
 
 
@@ -147,68 +147,68 @@ st.markdown(
 # ======================================================================
 
 with st.sidebar:
- st.markdown("## GraphMind")
- st.markdown(
- "<span class='accent-green'>Knowledge Graph</span> "
- "<span class='accent-blue'>Construction & Reasoning</span>",
- unsafe_allow_html=True,
- )
- st.markdown("---")
-
- # --- Input source ---
- st.markdown("### Text Source")
- input_mode = st.radio(
- "Choose input method",
- ["Demo Texts", "Paste Your Own"],
- label_visibility="collapsed",
- )
-
- text_to_process = ""
-
- if input_mode == "Demo Texts":
- selected_demo = st.selectbox(
- "Select a demo text",
- list(SAMPLE_TEXTS.keys()),
- )
- text_to_process = SAMPLE_TEXTS[selected_demo]
- with st.expander("Preview text", expanded=False):
- st.caption(text_to_process[:500] + "…")
- else:
- text_to_process = st.text_area(
- "Paste your text below",
- height=250,
- placeholder="Enter text containing named entities…",
- )
-
- st.markdown("---")
-
- # --- Extraction settings ---
- st.markdown("### Extraction Settings")
- entity_types = st.multiselect(
- "Entity types to extract",
- ["PERSON", "ORG", "LOCATION", "DATE", "TECHNOLOGY"],
- default=["PERSON", "ORG", "LOCATION", "DATE", "TECHNOLOGY"],
- )
-
- min_mentions = st.slider(
- "Minimum mentions for nodes",
- min_value=1,
- max_value=5,
- value=1,
- help="Only show entities mentioned at least this many times.",
- )
-
- st.markdown("---")
-
- # --- Build button ---
- build_clicked = st.button("Build Knowledge Graph", use_container_width=True)
-
- st.markdown("---")
- st.markdown(
- "<div style='text-align:center;color:#555;font-size:0.75rem;'>"
- "Built by <b>Yogesh Kuchimanchi</b><br>MIT License</div>",
- unsafe_allow_html=True,
- )
+    st.markdown("## GraphMind")
+    st.markdown(
+        "<span class='accent-green'>Knowledge Graph</span> "
+        "<span class='accent-blue'>Construction & Reasoning</span>",
+        unsafe_allow_html=True,
+    )
+    st.markdown("---")
+
+    # --- Input source ---
+    st.markdown("### Text Source")
+    input_mode = st.radio(
+        "Choose input method",
+        ["Demo Texts", "Paste Your Own"],
+        label_visibility="collapsed",
+    )
+
+    text_to_process = ""
+
+    if input_mode == "Demo Texts":
+        selected_demo = st.selectbox(
+            "Select a demo text",
+            list(SAMPLE_TEXTS.keys()),
+        )
+        text_to_process = SAMPLE_TEXTS[selected_demo]
+        with st.expander("Preview text", expanded=False):
+            st.caption(text_to_process[:500] + "…")
+    else:
+        text_to_process = st.text_area(
+            "Paste your text below",
+            height=250,
+            placeholder="Enter text containing named entities…",
+        )
+
+    st.markdown("---")
+
+    # --- Extraction settings ---
+    st.markdown("### Extraction Settings")
+    entity_types = st.multiselect(
+        "Entity types to extract",
+        ["PERSON", "ORG", "LOCATION", "DATE", "TECHNOLOGY"],
+        default=["PERSON", "ORG", "LOCATION", "DATE", "TECHNOLOGY"],
+    )
+
+    min_mentions = st.slider(
+        "Minimum mentions for nodes",
+        min_value=1,
+        max_value=5,
+        value=1,
+        help="Only show entities mentioned at least this many times.",
+    )
+
+    st.markdown("---")
+
+    # --- Build button ---
+    build_clicked = st.button(" Build Knowledge Graph", use_container_width=True)
+
+    st.markdown("---")
+    st.markdown(
+        "<div style='text-align:center;color:#555;font-size:0.75rem;'>"
+        "Built by <b>Yogesh Kuchimanchi</b><br>MIT License</div>",
+        unsafe_allow_html=True,
+    )
 
 
 # ======================================================================
@@ -216,25 +216,25 @@ with st.sidebar:
 # ======================================================================
 
 st.markdown(
- "<h1 style='text-align:center;'>"
- "Graph<span class='accent-green'>Mind</span></h1>",
- unsafe_allow_html=True,
+    "<h1 style='text-align:center;'>"
+    " Graph<span class='accent-green'>Mind</span></h1>",
+    unsafe_allow_html=True,
 )
 st.markdown(
- "<p style='text-align:center;color:#888;margin-top:-10px;'>"
- "Construct knowledge graphs from unstructured text using rule-based NER "
- "and graph reasoning.</p>",
- unsafe_allow_html=True,
+    "<p style='text-align:center;color:#888;margin-top:-10px;'>"
+    "Construct knowledge graphs from unstructured text using rule-based NER "
+    "and graph reasoning.</p>",
+    unsafe_allow_html=True,
 )
 
 # Colour legend
 legend_html = " ".join(
- f"<span class='legend-pill' style='background:{color};'>{label}</span>"
- for label, color in ENTITY_COLORS.items()
+    f"<span class='legend-pill' style='background:{color};'>{label}</span>"
+    for label, color in ENTITY_COLORS.items()
 )
 st.markdown(
- f"<div style='text-align:center;margin-bottom:20px;'>{legend_html}</div>",
- unsafe_allow_html=True,
+    f"<div style='text-align:center;margin-bottom:20px;'>{legend_html}</div>",
+    unsafe_allow_html=True,
 )
 
 
@@ -244,23 +244,23 @@ st.markdown(
 
 @st.cache_data(show_spinner=False)
 def run_pipeline(text: str, types: tuple):
- """Run NER + graph construction and cache results."""
- extractor = EntityExtractor()
- entities = extractor.extract(text)
+    """Run NER + graph construction and cache results."""
+    extractor = EntityExtractor()
+    entities = extractor.extract(text)
 
- # Filter entity types
- entities = [e for e in entities if e["label"] in types]
+    # Filter entity types
+    entities = [e for e in entities if e["label"] in types]
 
- relationships = extractor.extract_relationships(text, entities)
+    relationships = extractor.extract_relationships(text, entities)
 
- kg = KnowledgeGraph()
- kg.add_entities(entities)
- kg.add_relationships(relationships)
+    kg = KnowledgeGraph()
+    kg.add_entities(entities)
+    kg.add_relationships(relationships)
 
- stats = kg.get_stats()
- graph_html = create_pyvis_graph(kg)
+    stats = kg.get_stats()
+    graph_html = create_pyvis_graph(kg)
 
- return entities, relationships, kg, stats, graph_html
+    return entities, relationships, kg, stats, graph_html
 
 
 # ======================================================================
@@ -268,132 +268,132 @@ def run_pipeline(text: str, types: tuple):
 # ======================================================================
 
 if "has_run" not in st.session_state:
- st.session_state.has_run = False
+    st.session_state.has_run = False
 
 if build_clicked and text_to_process.strip():
- st.session_state.has_run = True
- st.session_state.text = text_to_process
- st.session_state.types = tuple(entity_types)
+    st.session_state.has_run = True
+    st.session_state.text = text_to_process
+    st.session_state.types = tuple(entity_types)
 
 # Auto-run on first visit with demo text
 if not st.session_state.has_run and input_mode == "Demo Texts":
- st.session_state.has_run = True
- st.session_state.text = text_to_process
- st.session_state.types = tuple(entity_types)
+    st.session_state.has_run = True
+    st.session_state.text = text_to_process
+    st.session_state.types = tuple(entity_types)
 
 if st.session_state.has_run:
- with st.spinner("Extracting entities and building graph…"):
- entities, relationships, kg, stats, graph_html = run_pipeline(
- st.session_state.text, st.session_state.types
- )
-
- # ==================================================================
- # Metrics row
- # ==================================================================
- m1, m2, m3, m4 = st.columns(4)
- m1.metric("Total Nodes", stats["total_nodes"])
- m2.metric("Total Edges", stats["total_edges"])
- m3.metric("Communities", stats["num_communities"])
- m4.metric("Entity Types", len(stats["entity_type_counts"]))
-
- st.markdown("---")
-
- # ==================================================================
- # Tabs
- # ==================================================================
- tab_graph, tab_entities, tab_relations, tab_stats = st.tabs(
- ["Interactive Graph", "Entities", "Relationships", "Statistics"]
- )
-
- # --- Interactive Graph ---
- with tab_graph:
- st.markdown("#### Interactive Knowledge Graph")
- st.caption("Drag, zoom, and hover nodes for details.")
- components.html(graph_html, height=680, scrolling=False)
-
- # --- Entities table ---
- with tab_entities:
- st.markdown("#### Extracted Entities")
- if entities:
- df_ent = pd.DataFrame(entities)
- df_ent = df_ent[["text", "label", "start", "end"]]
- df_ent.columns = ["Entity", "Type", "Start", "End"]
-
- # Colour-coded type column
- st.dataframe(
- df_ent.style.apply(
- lambda row: [
- "",
- f"color: {ENTITY_COLORS.get(row['Type'], '#888')}",
- "",
- "",
- ],
- axis=1,
- ),
- use_container_width=True,
- height=450,
- )
- st.caption(f"Total: **{len(entities)}** entities extracted.")
- else:
- st.info("No entities found. Try different text or settings.")
-
- # --- Relationships table ---
- with tab_relations:
- st.markdown("#### Extracted Relationships")
- if relationships:
- df_rel = pd.DataFrame(relationships)
- df_rel = df_rel[["source", "relation", "target", "source_label", "target_label"]]
- df_rel.columns = ["Source", "Relation", "Target", "Src Type", "Tgt Type"]
- st.dataframe(df_rel, use_container_width=True, height=450)
- st.caption(f"Total: **{len(relationships)}** relationships inferred.")
- else:
- st.info("No relationships found.")
-
- # --- Statistics ---
- with tab_stats:
- st.markdown("#### Graph Analytics")
-
- col_left, col_right = st.columns(2)
-
- with col_left:
- fig_dist = graph_stats_chart(stats)
- st.plotly_chart(fig_dist, use_container_width=True)
-
- with col_right:
- fig_community = community_chart(stats["communities"])
- st.plotly_chart(fig_community, use_container_width=True)
-
- st.markdown("---")
- fig_central = centrality_chart(stats["top_central_nodes"])
- st.plotly_chart(fig_central, use_container_width=True)
-
- with st.expander("Community Details"):
- for i, comm in enumerate(stats["communities"]):
- st.markdown(
- f"**Community {i+1}** ({len(comm)} members): "
- + ", ".join(comm)
- )
-
- with st.expander("Raw Statistics"):
- st.json(
- {
- "density": round(stats["density"], 6),
- "total_nodes": stats["total_nodes"],
- "total_edges": stats["total_edges"],
- "entity_type_counts": stats["entity_type_counts"],
- "relation_type_counts": stats["relation_type_counts"],
- "num_communities": stats["num_communities"],
- }
- )
+    with st.spinner("Extracting entities and building graph…"):
+        entities, relationships, kg, stats, graph_html = run_pipeline(
+            st.session_state.text, st.session_state.types
+        )
+
+    # ==================================================================
+    # Metrics row
+    # ==================================================================
+    m1, m2, m3, m4 = st.columns(4)
+    m1.metric("Total Nodes", stats["total_nodes"])
+    m2.metric("Total Edges", stats["total_edges"])
+    m3.metric("Communities", stats["num_communities"])
+    m4.metric("Entity Types", len(stats["entity_type_counts"]))
+
+    st.markdown("---")
+
+    # ==================================================================
+    # Tabs
+    # ==================================================================
+    tab_graph, tab_entities, tab_relations, tab_stats = st.tabs(
+        [" Interactive Graph", " Entities", " Relationships", " Statistics"]
+    )
+
+    # --- Interactive Graph ---
+    with tab_graph:
+        st.markdown("#### Interactive Knowledge Graph")
+        st.caption("Drag, zoom, and hover nodes for details.")
+        components.html(graph_html, height=680, scrolling=False)
+
+    # --- Entities table ---
+    with tab_entities:
+        st.markdown("#### Extracted Entities")
+        if entities:
+            df_ent = pd.DataFrame(entities)
+            df_ent = df_ent[["text", "label", "start", "end"]]
+            df_ent.columns = ["Entity", "Type", "Start", "End"]
+
+            # Colour-coded type column
+            st.dataframe(
+                df_ent.style.apply(
+                    lambda row: [
+                        "",
+                        f"color: {ENTITY_COLORS.get(row['Type'], '#888')}",
+                        "",
+                        "",
+                    ],
+                    axis=1,
+                ),
+                use_container_width=True,
+                height=450,
+            )
+            st.caption(f"Total: **{len(entities)}** entities extracted.")
+        else:
+            st.info("No entities found. Try different text or settings.")
+
+    # --- Relationships table ---
+    with tab_relations:
+        st.markdown("#### Extracted Relationships")
+        if relationships:
+            df_rel = pd.DataFrame(relationships)
+            df_rel = df_rel[["source", "relation", "target", "source_label", "target_label"]]
+            df_rel.columns = ["Source", "Relation", "Target", "Src Type", "Tgt Type"]
+            st.dataframe(df_rel, use_container_width=True, height=450)
+            st.caption(f"Total: **{len(relationships)}** relationships inferred.")
+        else:
+            st.info("No relationships found.")
+
+    # --- Statistics ---
+    with tab_stats:
+        st.markdown("#### Graph Analytics")
+
+        col_left, col_right = st.columns(2)
+
+        with col_left:
+            fig_dist = graph_stats_chart(stats)
+            st.plotly_chart(fig_dist, use_container_width=True)
+
+        with col_right:
+            fig_community = community_chart(stats["communities"])
+            st.plotly_chart(fig_community, use_container_width=True)
+
+        st.markdown("---")
+        fig_central = centrality_chart(stats["top_central_nodes"])
+        st.plotly_chart(fig_central, use_container_width=True)
+
+        with st.expander("Community Details"):
+            for i, comm in enumerate(stats["communities"]):
+                st.markdown(
+                    f"**Community {i+1}** ({len(comm)} members): "
+                    + ", ".join(comm)
+                )
+
+        with st.expander("Raw Statistics"):
+            st.json(
+                {
+                    "density": round(stats["density"], 6),
+                    "total_nodes": stats["total_nodes"],
+                    "total_edges": stats["total_edges"],
+                    "entity_type_counts": stats["entity_type_counts"],
+                    "relation_type_counts": stats["relation_type_counts"],
+                    "num_communities": stats["num_communities"],
+                }
+            )
 
 else:
- # Placeholder when nothing has been processed yet
- st.markdown(
- "<div style='text-align:center;padding:80px 20px;color:#555;'>"
- "<h3>Paste text or select a demo, then click "
- "<span class='accent-green'>Build Knowledge Graph</span></h3>"
- "<p>The engine will extract entities, infer relationships, "
- "and visualise an interactive knowledge graph.</p>"
- "</div>",
- unsafe_allow_html=True,
- )
+    # Placeholder when nothing has been processed yet
+    st.markdown(
+        "<div style='text-align:center;padding:80px 20px;color:#555;'>"
+        "<h3>Paste text or select a demo, then click "
+        "<span class='accent-green'>Build Knowledge Graph</span></h3>"
+        "<p>The engine will extract entities, infer relationships, "
+        "and visualise an interactive knowledge graph.</p>"
+        "</div>",
+        unsafe_allow_html=True,
+    )

src/extractor.py

@@ -15,250 +15,250 @@ from typing import List, Dict, Tuple
 # ---------------------------------------------------------------------------
 
 PERSON_PATTERNS = [
- # Titles followed by capitalized names
- r"(?:Dr|Prof|Mr|Mrs|Ms|Sir|Lord|President|CEO|CTO|Director)\.\s+[A-Z][a-z]+(?:\s+[A-Z][a-z]+)+",
- # Common well-known names (seed list)
- r"\b(?:Elon Musk|Jeff Bezos|Sam Altman|Demis Hassabis|Yann LeCun|Geoffrey Hinton|"
- r"Fei-Fei Li|Andrew Ng|Ilya Sutskever|Jensen Huang|Satya Nadella|Tim Cook|"
- r"Mark Zuckerberg|Sundar Pichai|Dario Amodei|Andrej Karpathy|"
- r"Alan Turing|Ada Lovelace|John von Neumann|Claude Shannon|"
- r"Albert Einstein|Isaac Newton|Marie Curie|Nikola Tesla|"
- r"Napoleon Bonaparte|Winston Churchill|Abraham Lincoln|Mahatma Gandhi|"
- r"Alexander Hamilton|Thomas Jefferson|Benjamin Franklin|George Washington|"
- r"Leonardo da Vinci|Galileo Galilei|Charles Darwin|Stephen Hawking)\b",
- # Two or three capitalized words that look like person names
- r"\b[A-Z][a-z]{2,15}\s+(?:[A-Z]\.\s+)?[A-Z][a-z]{2,15}\b",
+    # Titles followed by capitalized names
+    r"(?:Dr|Prof|Mr|Mrs|Ms|Sir|Lord|President|CEO|CTO|Director)\.\s+[A-Z][a-z]+(?:\s+[A-Z][a-z]+)+",
+    # Common well-known names (seed list)
+    r"\b(?:Elon Musk|Jeff Bezos|Sam Altman|Demis Hassabis|Yann LeCun|Geoffrey Hinton|"
+    r"Fei-Fei Li|Andrew Ng|Ilya Sutskever|Jensen Huang|Satya Nadella|Tim Cook|"
+    r"Mark Zuckerberg|Sundar Pichai|Dario Amodei|Andrej Karpathy|"
+    r"Alan Turing|Ada Lovelace|John von Neumann|Claude Shannon|"
+    r"Albert Einstein|Isaac Newton|Marie Curie|Nikola Tesla|"
+    r"Napoleon Bonaparte|Winston Churchill|Abraham Lincoln|Mahatma Gandhi|"
+    r"Alexander Hamilton|Thomas Jefferson|Benjamin Franklin|George Washington|"
+    r"Leonardo da Vinci|Galileo Galilei|Charles Darwin|Stephen Hawking)\b",
+    # Two or three capitalized words that look like person names
+    r"\b[A-Z][a-z]{2,15}\s+(?:[A-Z]\.\s+)?[A-Z][a-z]{2,15}\b",
 ]
 
 ORG_PATTERNS = [
- r"\b(?:Google|Microsoft|Apple|Amazon|Meta|OpenAI|DeepMind|Anthropic|Tesla|"
- r"NVIDIA|IBM|Intel|AMD|Qualcomm|Samsung|TSMC|Oracle|Salesforce|Adobe|"
- r"Netflix|Spotify|Twitter|LinkedIn|GitHub|Stack Overflow|"
- r"MIT|Stanford|Harvard|Oxford|Cambridge|Berkeley|Carnegie Mellon|"
- r"NASA|CERN|WHO|UNESCO|United Nations|European Union|"
- r"IEEE|ACM|NeurIPS|ICML|ICLR|AAAI|CVPR|"
- r"Goldman Sachs|JPMorgan|Morgan Stanley|BlackRock)\b",
- r"\b[A-Z][a-z]+(?:\s+[A-Z][a-z]+)*\s+(?:Inc|Corp|Ltd|LLC|Group|Foundation|"
- r"Institute|University|Laboratory|Labs|Research|Association|Organization)\b",
- r"\b(?:University|Institute|Academy)\s+of\s+[A-Z][a-z]+(?:\s+[A-Z][a-z]+)*\b",
+    r"\b(?:Google|Microsoft|Apple|Amazon|Meta|OpenAI|DeepMind|Anthropic|Tesla|"
+    r"NVIDIA|IBM|Intel|AMD|Qualcomm|Samsung|TSMC|Oracle|Salesforce|Adobe|"
+    r"Netflix|Spotify|Twitter|LinkedIn|GitHub|Stack Overflow|"
+    r"MIT|Stanford|Harvard|Oxford|Cambridge|Berkeley|Carnegie Mellon|"
+    r"NASA|CERN|WHO|UNESCO|United Nations|European Union|"
+    r"IEEE|ACM|NeurIPS|ICML|ICLR|AAAI|CVPR|"
+    r"Goldman Sachs|JPMorgan|Morgan Stanley|BlackRock)\b",
+    r"\b[A-Z][a-z]+(?:\s+[A-Z][a-z]+)*\s+(?:Inc|Corp|Ltd|LLC|Group|Foundation|"
+    r"Institute|University|Laboratory|Labs|Research|Association|Organization)\b",
+    r"\b(?:University|Institute|Academy)\s+of\s+[A-Z][a-z]+(?:\s+[A-Z][a-z]+)*\b",
 ]
 
 LOCATION_PATTERNS = [
- r"\b(?:New York|San Francisco|Silicon Valley|Los Angeles|Chicago|Boston|Seattle|"
- r"Washington D\.C\.|London|Paris|Berlin|Tokyo|Beijing|Shanghai|Mumbai|"
- r"Bangalore|Toronto|Montreal|Sydney|Singapore|Hong Kong|Dubai|"
- r"California|Texas|Massachusetts|Virginia|"
- r"United States|United Kingdom|China|India|Japan|Germany|France|Canada|"
- r"Australia|South Korea|Israel|Switzerland|"
- r"Europe|Asia|North America|South America|Africa)\b",
+    r"\b(?:New York|San Francisco|Silicon Valley|Los Angeles|Chicago|Boston|Seattle|"
+    r"Washington D\.C\.|London|Paris|Berlin|Tokyo|Beijing|Shanghai|Mumbai|"
+    r"Bangalore|Toronto|Montreal|Sydney|Singapore|Hong Kong|Dubai|"
+    r"California|Texas|Massachusetts|Virginia|"
+    r"United States|United Kingdom|China|India|Japan|Germany|France|Canada|"
+    r"Australia|South Korea|Israel|Switzerland|"
+    r"Europe|Asia|North America|South America|Africa)\b",
 ]
 
 DATE_PATTERNS = [
- # Full dates
- r"\b(?:January|February|March|April|May|June|July|August|September|"
- r"October|November|December)\s+\d{1,2},?\s+\d{4}\b",
- # Month Year
- r"\b(?:January|February|March|April|May|June|July|August|September|"
- r"October|November|December)\s+\d{4}\b",
- # Year ranges & standalone years
- r"\b(?:19|20)\d{2}[-–]\d{2,4}\b",
- r"\b(?:19|20)\d{2}s?\b",
- # Relative dates
- r"\b(?:Q[1-4]\s+\d{4})\b",
+    # Full dates
+    r"\b(?:January|February|March|April|May|June|July|August|September|"
+    r"October|November|December)\s+\d{1,2},?\s+\d{4}\b",
+    # Month Year
+    r"\b(?:January|February|March|April|May|June|July|August|September|"
+    r"October|November|December)\s+\d{4}\b",
+    # Year ranges & standalone years
+    r"\b(?:19|20)\d{2}[-–]\d{2,4}\b",
+    r"\b(?:19|20)\d{2}s?\b",
+    # Relative dates
+    r"\b(?:Q[1-4]\s+\d{4})\b",
 ]
 
 TECHNOLOGY_PATTERNS = [
- r"\b(?:GPT-[0-9]+|GPT|BERT|Transformer|LLM|LLMs|DALL[-·]E|Stable Diffusion|"
- r"ChatGPT|Copilot|AlphaFold|AlphaGo|"
- r"Python|JavaScript|TypeScript|Rust|Go|Java|C\+\+|SQL|"
- r"TensorFlow|PyTorch|Keras|scikit-learn|Hugging Face|LangChain|"
- r"Kubernetes|Docker|AWS|Azure|GCP|"
- r"blockchain|quantum computing|machine learning|deep learning|"
- r"artificial intelligence|natural language processing|NLP|"
- r"computer vision|reinforcement learning|neural network|neural networks|"
- r"convolutional neural network|CNN|RNN|LSTM|GAN|GANs|"
- r"large language model|retrieval-augmented generation|RAG|"
- r"knowledge graph|attention mechanism|self-attention)\b",
+    r"\b(?:GPT-[0-9]+|GPT|BERT|Transformer|LLM|LLMs|DALL[-·]E|Stable Diffusion|"
+    r"ChatGPT|Copilot|AlphaFold|AlphaGo|"
+    r"Python|JavaScript|TypeScript|Rust|Go|Java|C\+\+|SQL|"
+    r"TensorFlow|PyTorch|Keras|scikit-learn|Hugging Face|LangChain|"
+    r"Kubernetes|Docker|AWS|Azure|GCP|"
+    r"blockchain|quantum computing|machine learning|deep learning|"
+    r"artificial intelligence|natural language processing|NLP|"
+    r"computer vision|reinforcement learning|neural network|neural networks|"
+    r"convolutional neural network|CNN|RNN|LSTM|GAN|GANs|"
+    r"large language model|retrieval-augmented generation|RAG|"
+    r"knowledge graph|attention mechanism|self-attention)\b",
 ]
 
 # Map label -> compiled patterns
 ENTITY_PATTERNS: Dict[str, List[re.Pattern]] = {
- "TECHNOLOGY": [re.compile(p, re.IGNORECASE) for p in TECHNOLOGY_PATTERNS],
- "ORG": [re.compile(p) for p in ORG_PATTERNS],
- "LOCATION": [re.compile(p) for p in LOCATION_PATTERNS],
- "DATE": [re.compile(p) for p in DATE_PATTERNS],
- "PERSON": [re.compile(p) for p in PERSON_PATTERNS],
+    "TECHNOLOGY": [re.compile(p, re.IGNORECASE) for p in TECHNOLOGY_PATTERNS],
+    "ORG":        [re.compile(p) for p in ORG_PATTERNS],
+    "LOCATION":   [re.compile(p) for p in LOCATION_PATTERNS],
+    "DATE":       [re.compile(p) for p in DATE_PATTERNS],
+    "PERSON":     [re.compile(p) for p in PERSON_PATTERNS],
 }
 
 # Words that should never be tagged as PERSON
 PERSON_STOPWORDS = {
- "The", "This", "That", "These", "Those", "Here", "There",
- "However", "Moreover", "Furthermore", "Although", "Because",
- "While", "During", "After", "Before", "Since", "Within",
- "Between", "Through", "About", "Their", "Where", "Which",
- "Every", "Other", "Another", "First", "Second", "Third",
- "Many", "Most", "Some", "Such", "Each", "Both", "Several",
- "Recent", "Major", "Large", "Small", "High", "Early", "Late",
- "With", "From", "Into", "Over", "Under", "Also", "Just",
- "More", "Very", "Much", "Well", "Even", "Still", "Already",
- "Knowledge Graph", "Construction", "Reasoning", "Engine",
- "Research", "Development", "Analysis", "Processing", "Learning",
+    "The", "This", "That", "These", "Those", "Here", "There",
+    "However", "Moreover", "Furthermore", "Although", "Because",
+    "While", "During", "After", "Before", "Since", "Within",
+    "Between", "Through", "About", "Their", "Where", "Which",
+    "Every", "Other", "Another", "First", "Second", "Third",
+    "Many", "Most", "Some", "Such", "Each", "Both", "Several",
+    "Recent", "Major", "Large", "Small", "High", "Early", "Late",
+    "With", "From", "Into", "Over", "Under", "Also", "Just",
+    "More", "Very", "Much", "Well", "Even", "Still", "Already",
+    "Knowledge Graph", "Construction", "Reasoning", "Engine",
+    "Research", "Development", "Analysis", "Processing", "Learning",
 }
 
 
 class EntityExtractor:
- """
- Rule-based Named Entity Recognition engine.
-
- Uses curated regex patterns to identify entities in text without
- requiring large spaCy model downloads.
- """
-
- def __init__(self):
- self.patterns = ENTITY_PATTERNS
-
- # ------------------------------------------------------------------
- # Public API
- # ------------------------------------------------------------------
-
- def extract(self, text: str) -> List[Dict]:
- """
- Extract named entities from *text*.
-
- Returns a list of dicts:
- [{"text": ..., "label": ..., "start": ..., "end": ...}, ...]
- """
- raw_entities: List[Dict] = []
-
- for label, compiled_patterns in self.patterns.items():
- for pattern in compiled_patterns:
- for match in pattern.finditer(text):
- entity_text = match.group().strip()
-
- # Filter noisy PERSON matches
- if label == "PERSON" and entity_text in PERSON_STOPWORDS:
- continue
- if label == "PERSON" and len(entity_text.split()) < 2:
- continue
-
- raw_entities.append({
- "text": entity_text,
- "label": label,
- "start": match.start(),
- "end": match.end(),
- })
-
- # Deduplicate overlapping spans (prefer longer matches)
- entities = self._resolve_overlaps(raw_entities)
- return entities
-
- def extract_relationships(
- self, text: str, entities: List[Dict] | None = None
- ) -> List[Dict]:
- """
- Infer relationships between entities via sentence co-occurrence.
-
- Returns a list of dicts:
- [{"source": ..., "target": ..., "relation": ..., "sentence": ...}, ...]
- """
- if entities is None:
- entities = self.extract(text)
-
- sentences = self._split_sentences(text)
- relationships: List[Dict] = []
- seen: set = set()
-
- for sentence in sentences:
- # Find entities present in this sentence
- present = [
- e for e in entities
- if e["text"] in sentence
- ]
-
- for i, src in enumerate(present):
- for tgt in present[i + 1:]:
- key = (src["text"], tgt["text"])
- if key in seen:
- continue
- seen.add(key)
-
- relation = self._infer_relation(src, tgt, sentence)
- relationships.append({
- "source": src["text"],
- "target": tgt["text"],
- "source_label": src["label"],
- "target_label": tgt["label"],
- "relation": relation,
- "sentence": sentence.strip(),
- })
-
- return relationships
-
- # ------------------------------------------------------------------
- # Internal helpers
- # ------------------------------------------------------------------
-
- @staticmethod
- def _resolve_overlaps(entities: List[Dict]) -> List[Dict]:
- """Keep the longest span when two entities overlap."""
- # Sort by start, then by descending length
- entities.sort(key=lambda e: (e["start"], -(e["end"] - e["start"])))
- result: List[Dict] = []
- last_end = -1
- for ent in entities:
- if ent["start"] >= last_end:
- result.append(ent)
- last_end = ent["end"]
- return result
-
- @staticmethod
- def _split_sentences(text: str) -> List[str]:
- """Naive sentence splitter."""
- return re.split(r"(?<=[.!?])\s+", text)
-
- @staticmethod
- def _infer_relation(src: Dict, tgt: Dict, sentence: str) -> str:
- """Heuristic relation labelling based on entity types and context."""
- pair = (src["label"], tgt["label"])
-
- # Keyword-based relation detection
- s_lower = sentence.lower()
-
- if any(kw in s_lower for kw in ["founded", "co-founded", "started", "created"]):
- if pair in [("PERSON", "ORG"), ("PERSON", "TECHNOLOGY")]:
- return "FOUNDED"
- if any(kw in s_lower for kw in ["acquired", "bought", "purchased", "merged"]):
- return "ACQUIRED"
- if any(kw in s_lower for kw in ["works at", "joined", "hired", "employed"]):
- return "WORKS_AT"
- if any(kw in s_lower for kw in ["located in", "based in", "headquartered"]):
- return "LOCATED_IN"
- if any(kw in s_lower for kw in ["developed", "built", "designed", "invented"]):
- return "DEVELOPED"
- if any(kw in s_lower for kw in ["published", "released", "announced", "launched"]):
- return "RELEASED"
- if any(kw in s_lower for kw in ["uses", "using", "powered by", "built on", "leverages"]):
- return "USES"
- if any(kw in s_lower for kw in ["competed", "versus", "rivaling", "competing"]):
- return "COMPETES_WITH"
- if any(kw in s_lower for kw in ["collaborated", "partnered", "partnership"]):
- return "COLLABORATES_WITH"
- if any(kw in s_lower for kw in ["invested", "funding", "backed"]):
- return "INVESTED_IN"
-
- # Fallback: type-pair heuristics
- relation_map = {
- ("PERSON", "ORG"): "AFFILIATED_WITH",
- ("PERSON", "TECHNOLOGY"): "WORKS_ON",
- ("PERSON", "LOCATION"): "LOCATED_IN",
- ("ORG", "TECHNOLOGY"): "DEVELOPS",
- ("ORG", "LOCATION"): "LOCATED_IN",
- ("ORG", "ORG"): "RELATED_TO",
- ("TECHNOLOGY", "TECHNOLOGY"): "RELATED_TO",
- ("PERSON", "PERSON"): "ASSOCIATED_WITH",
- ("PERSON", "DATE"): "ACTIVE_IN",
- ("ORG", "DATE"): "ACTIVE_IN",
- ("TECHNOLOGY", "DATE"): "EMERGED_IN",
- }
-
- return relation_map.get(pair, relation_map.get((tgt["label"], src["label"]), "RELATED_TO"))
+    """
+    Rule-based Named Entity Recognition engine.
+
+    Uses curated regex patterns to identify entities in text without
+    requiring large spaCy model downloads.
+    """
+
+    def __init__(self):
+        self.patterns = ENTITY_PATTERNS
+
+    # ------------------------------------------------------------------
+    # Public API
+    # ------------------------------------------------------------------
+
+    def extract(self, text: str) -> List[Dict]:
+        """
+        Extract named entities from *text*.
+
+        Returns a list of dicts:
+            [{"text": ..., "label": ..., "start": ..., "end": ...}, ...]
+        """
+        raw_entities: List[Dict] = []
+
+        for label, compiled_patterns in self.patterns.items():
+            for pattern in compiled_patterns:
+                for match in pattern.finditer(text):
+                    entity_text = match.group().strip()
+
+                    # Filter noisy PERSON matches
+                    if label == "PERSON" and entity_text in PERSON_STOPWORDS:
+                        continue
+                    if label == "PERSON" and len(entity_text.split()) < 2:
+                        continue
+
+                    raw_entities.append({
+                        "text": entity_text,
+                        "label": label,
+                        "start": match.start(),
+                        "end": match.end(),
+                    })
+
+        # Deduplicate overlapping spans (prefer longer matches)
+        entities = self._resolve_overlaps(raw_entities)
+        return entities
+
+    def extract_relationships(
+        self, text: str, entities: List[Dict] | None = None
+    ) -> List[Dict]:
+        """
+        Infer relationships between entities via sentence co-occurrence.
+
+        Returns a list of dicts:
+            [{"source": ..., "target": ..., "relation": ..., "sentence": ...}, ...]
+        """
+        if entities is None:
+            entities = self.extract(text)
+
+        sentences = self._split_sentences(text)
+        relationships: List[Dict] = []
+        seen: set = set()
+
+        for sentence in sentences:
+            # Find entities present in this sentence
+            present = [
+                e for e in entities
+                if e["text"] in sentence
+            ]
+
+            for i, src in enumerate(present):
+                for tgt in present[i + 1:]:
+                    key = (src["text"], tgt["text"])
+                    if key in seen:
+                        continue
+                    seen.add(key)
+
+                    relation = self._infer_relation(src, tgt, sentence)
+                    relationships.append({
+                        "source": src["text"],
+                        "target": tgt["text"],
+                        "source_label": src["label"],
+                        "target_label": tgt["label"],
+                        "relation": relation,
+                        "sentence": sentence.strip(),
+                    })
+
+        return relationships
+
+    # ------------------------------------------------------------------
+    # Internal helpers
+    # ------------------------------------------------------------------
+
+    @staticmethod
+    def _resolve_overlaps(entities: List[Dict]) -> List[Dict]:
+        """Keep the longest span when two entities overlap."""
+        # Sort by start, then by descending length
+        entities.sort(key=lambda e: (e["start"], -(e["end"] - e["start"])))
+        result: List[Dict] = []
+        last_end = -1
+        for ent in entities:
+            if ent["start"] >= last_end:
+                result.append(ent)
+                last_end = ent["end"]
+        return result
+
+    @staticmethod
+    def _split_sentences(text: str) -> List[str]:
+        """Naive sentence splitter."""
+        return re.split(r"(?<=[.!?])\s+", text)
+
+    @staticmethod
+    def _infer_relation(src: Dict, tgt: Dict, sentence: str) -> str:
+        """Heuristic relation labelling based on entity types and context."""
+        pair = (src["label"], tgt["label"])
+
+        # Keyword-based relation detection
+        s_lower = sentence.lower()
+
+        if any(kw in s_lower for kw in ["founded", "co-founded", "started", "created"]):
+            if pair in [("PERSON", "ORG"), ("PERSON", "TECHNOLOGY")]:
+                return "FOUNDED"
+        if any(kw in s_lower for kw in ["acquired", "bought", "purchased", "merged"]):
+            return "ACQUIRED"
+        if any(kw in s_lower for kw in ["works at", "joined", "hired", "employed"]):
+            return "WORKS_AT"
+        if any(kw in s_lower for kw in ["located in", "based in", "headquartered"]):
+            return "LOCATED_IN"
+        if any(kw in s_lower for kw in ["developed", "built", "designed", "invented"]):
+            return "DEVELOPED"
+        if any(kw in s_lower for kw in ["published", "released", "announced", "launched"]):
+            return "RELEASED"
+        if any(kw in s_lower for kw in ["uses", "using", "powered by", "built on", "leverages"]):
+            return "USES"
+        if any(kw in s_lower for kw in ["competed", "versus", "rivaling", "competing"]):
+            return "COMPETES_WITH"
+        if any(kw in s_lower for kw in ["collaborated", "partnered", "partnership"]):
+            return "COLLABORATES_WITH"
+        if any(kw in s_lower for kw in ["invested", "funding", "backed"]):
+            return "INVESTED_IN"
+
+        # Fallback: type-pair heuristics
+        relation_map = {
+            ("PERSON", "ORG"): "AFFILIATED_WITH",
+            ("PERSON", "TECHNOLOGY"): "WORKS_ON",
+            ("PERSON", "LOCATION"): "LOCATED_IN",
+            ("ORG", "TECHNOLOGY"): "DEVELOPS",
+            ("ORG", "LOCATION"): "LOCATED_IN",
+            ("ORG", "ORG"): "RELATED_TO",
+            ("TECHNOLOGY", "TECHNOLOGY"): "RELATED_TO",
+            ("PERSON", "PERSON"): "ASSOCIATED_WITH",
+            ("PERSON", "DATE"): "ACTIVE_IN",
+            ("ORG", "DATE"): "ACTIVE_IN",
+            ("TECHNOLOGY", "DATE"): "EMERGED_IN",
+        }
+
+        return relation_map.get(pair, relation_map.get((tgt["label"], src["label"]), "RELATED_TO"))

src/graph_builder.py

@@ -13,232 +13,232 @@ import json
 
 # Colour palette for entity types
 ENTITY_COLORS = {
- "PERSON": "#00ff88",
- "ORG": "#00d4ff",
- "LOCATION": "#a855f7",
- "TECHNOLOGY": "#f59e0b",
- "DATE": "#ec4899",
+    "PERSON": "#00ff88",
+    "ORG": "#00d4ff",
+    "LOCATION": "#a855f7",
+    "TECHNOLOGY": "#f59e0b",
+    "DATE": "#ec4899",
 }
 
 ENTITY_SHAPES = {
- "PERSON": "dot",
- "ORG": "diamond",
- "LOCATION": "triangle",
- "TECHNOLOGY": "square",
- "DATE": "star",
+    "PERSON": "dot",
+    "ORG": "diamond",
+    "LOCATION": "triangle",
+    "TECHNOLOGY": "square",
+    "DATE": "star",
 }
 
 
 class KnowledgeGraph:
- """
- A directed knowledge graph backed by ``networkx.DiGraph``.
- """
-
- def __init__(self):
- self.graph = nx.DiGraph()
-
- # ------------------------------------------------------------------
- # Construction
- # ------------------------------------------------------------------
-
- def add_entities(self, entities: List[Dict]) -> None:
- """
- Add entity nodes to the graph.
-
- Parameters
- ----------
- entities : list of dict
- Each dict must contain at least ``text`` and ``label`` keys.
- """
- for ent in entities:
- node_id = ent["text"]
- if self.graph.has_node(node_id):
- # Increment mention count
- self.graph.nodes[node_id]["mentions"] = (
- self.graph.nodes[node_id].get("mentions", 1) + 1
- )
- continue
-
- self.graph.add_node(
- node_id,
- label=ent["label"],
- color=ENTITY_COLORS.get(ent["label"], "#888888"),
- shape=ENTITY_SHAPES.get(ent["label"], "dot"),
- mentions=1,
- )
-
- def add_relationships(self, relationships: List[Dict]) -> None:
- """
- Add directed edges (relationships) to the graph.
-
- Parameters
- ----------
- relationships : list of dict
- Each dict needs ``source``, ``target``, ``relation`` keys.
- """
- for rel in relationships:
- src, tgt = rel["source"], rel["target"]
-
- # Ensure nodes exist
- if not self.graph.has_node(src):
- self.graph.add_node(
- src,
- label=rel.get("source_label", "UNKNOWN"),
- color=ENTITY_COLORS.get(rel.get("source_label"), "#888888"),
- shape=ENTITY_SHAPES.get(rel.get("source_label"), "dot"),
- mentions=1,
- )
- if not self.graph.has_node(tgt):
- self.graph.add_node(
- tgt,
- label=rel.get("target_label", "UNKNOWN"),
- color=ENTITY_COLORS.get(rel.get("target_label"), "#888888"),
- shape=ENTITY_SHAPES.get(rel.get("target_label"), "dot"),
- mentions=1,
- )
-
- if self.graph.has_edge(src, tgt):
- self.graph.edges[src, tgt]["weight"] = (
- self.graph.edges[src, tgt].get("weight", 1) + 1
- )
- else:
- self.graph.add_edge(
- src,
- tgt,
- relation=rel["relation"],
- weight=1,
- sentence=rel.get("sentence", ""),
- )
-
- # ------------------------------------------------------------------
- # Queries & Analytics
- # ------------------------------------------------------------------
-
- def get_stats(self) -> Dict[str, Any]:
- """Return summary statistics of the knowledge graph."""
- G = self.graph
-
- label_counts: Dict[str, int] = {}
- for _, data in G.nodes(data=True):
- lbl = data.get("label", "UNKNOWN")
- label_counts[lbl] = label_counts.get(lbl, 0) + 1
-
- relation_counts: Dict[str, int] = {}
- for _, _, data in G.edges(data=True):
- rel = data.get("relation", "UNKNOWN")
- relation_counts[rel] = relation_counts.get(rel, 0) + 1
-
- communities = self.get_communities()
-
- # Degree centrality for top nodes
- if G.number_of_nodes() > 0:
- centrality = nx.degree_centrality(G)
- top_nodes = sorted(centrality.items(), key=lambda x: x[1], reverse=True)[:10]
- else:
- top_nodes = []
-
- return {
- "total_nodes": G.number_of_nodes(),
- "total_edges": G.number_of_edges(),
- "entity_type_counts": label_counts,
- "relation_type_counts": relation_counts,
- "num_communities": len(communities),
- "communities": communities,
- "top_central_nodes": top_nodes,
- "density": nx.density(G) if G.number_of_nodes() > 1 else 0,
- }
-
- def get_communities(self) -> List[List[str]]:
- """
- Detect communities using the greedy modularity algorithm
- on the undirected projection.
- """
- if self.graph.number_of_nodes() == 0:
- return []
-
- undirected = self.graph.to_undirected()
- try:
- from networkx.algorithms.community import greedy_modularity_communities
- communities = greedy_modularity_communities(undirected)
- return [sorted(list(c)) for c in communities]
- except Exception:
- # Fallback: connected components
- return [sorted(list(c)) for c in nx.connected_components(undirected)]
-
- def get_node_details(self, node_id: str) -> Optional[Dict]:
- """Return all attributes for a single node."""
- if not self.graph.has_node(node_id):
- return None
- data = dict(self.graph.nodes[node_id])
- data["id"] = node_id
- data["in_degree"] = self.graph.in_degree(node_id)
- data["out_degree"] = self.graph.out_degree(node_id)
- data["neighbors"] = list(self.graph.successors(node_id)) + list(
- self.graph.predecessors(node_id)
- )
- return data
-
- # ------------------------------------------------------------------
- # Export
- # ------------------------------------------------------------------
-
- def to_pyvis(self, height: str = "600px", width: str = "100%") -> Network:
- """
- Convert the graph to a PyVis ``Network`` for interactive
- HTML visualisation.
- """
- net = Network(
- height=height,
- width=width,
- directed=True,
- bgcolor="#0a0a0a",
- font_color="white",
- select_menu=False,
- filter_menu=False,
- )
-
- net.barnes_hut(
- gravity=-8000,
- central_gravity=0.3,
- spring_length=200,
- spring_strength=0.05,
- damping=0.09,
- )
-
- for node_id, data in self.graph.nodes(data=True):
- mentions = data.get("mentions", 1)
- size = 15 + mentions * 5
- net.add_node(
- node_id,
- label=node_id,
- color=data.get("color", "#888888"),
- shape=data.get("shape", "dot"),
- size=min(size, 50),
- title=f"{data.get('label', 'UNKNOWN')}\nMentions: {mentions}",
- font={"size": 14, "color": "white"},
- )
-
- for src, tgt, data in self.graph.edges(data=True):
- relation = data.get("relation", "")
- weight = data.get("weight", 1)
- net.add_edge(
- src,
- tgt,
- title=relation,
- label=relation,
- width=min(weight * 1.5, 6),
- color={"color": "#444444", "highlight": "#00ff88"},
- font={"size": 10, "color": "#888888", "align": "middle"},
- arrows={"to": {"enabled": True, "scaleFactor": 0.5}},
- smooth={"type": "curvedCW", "roundness": 0.2},
- )
-
- return net
-
- def to_dict(self) -> Dict:
- """Serialise the graph to a JSON-safe dictionary."""
- return nx.node_link_data(self.graph)
-
- def from_dict(self, data: Dict) -> None:
- """Load a graph from a dictionary produced by ``to_dict``."""
- self.graph = nx.node_link_graph(data)
+    """
+    A directed knowledge graph backed by ``networkx.DiGraph``.
+    """
+
+    def __init__(self):
+        self.graph = nx.DiGraph()
+
+    # ------------------------------------------------------------------
+    # Construction
+    # ------------------------------------------------------------------
+
+    def add_entities(self, entities: List[Dict]) -> None:
+        """
+        Add entity nodes to the graph.
+
+        Parameters
+        ----------
+        entities : list of dict
+            Each dict must contain at least ``text`` and ``label`` keys.
+        """
+        for ent in entities:
+            node_id = ent["text"]
+            if self.graph.has_node(node_id):
+                # Increment mention count
+                self.graph.nodes[node_id]["mentions"] = (
+                    self.graph.nodes[node_id].get("mentions", 1) + 1
+                )
+                continue
+
+            self.graph.add_node(
+                node_id,
+                label=ent["label"],
+                color=ENTITY_COLORS.get(ent["label"], "#888888"),
+                shape=ENTITY_SHAPES.get(ent["label"], "dot"),
+                mentions=1,
+            )
+
+    def add_relationships(self, relationships: List[Dict]) -> None:
+        """
+        Add directed edges (relationships) to the graph.
+
+        Parameters
+        ----------
+        relationships : list of dict
+            Each dict needs ``source``, ``target``, ``relation`` keys.
+        """
+        for rel in relationships:
+            src, tgt = rel["source"], rel["target"]
+
+            # Ensure nodes exist
+            if not self.graph.has_node(src):
+                self.graph.add_node(
+                    src,
+                    label=rel.get("source_label", "UNKNOWN"),
+                    color=ENTITY_COLORS.get(rel.get("source_label"), "#888888"),
+                    shape=ENTITY_SHAPES.get(rel.get("source_label"), "dot"),
+                    mentions=1,
+                )
+            if not self.graph.has_node(tgt):
+                self.graph.add_node(
+                    tgt,
+                    label=rel.get("target_label", "UNKNOWN"),
+                    color=ENTITY_COLORS.get(rel.get("target_label"), "#888888"),
+                    shape=ENTITY_SHAPES.get(rel.get("target_label"), "dot"),
+                    mentions=1,
+                )
+
+            if self.graph.has_edge(src, tgt):
+                self.graph.edges[src, tgt]["weight"] = (
+                    self.graph.edges[src, tgt].get("weight", 1) + 1
+                )
+            else:
+                self.graph.add_edge(
+                    src,
+                    tgt,
+                    relation=rel["relation"],
+                    weight=1,
+                    sentence=rel.get("sentence", ""),
+                )
+
+    # ------------------------------------------------------------------
+    # Queries & Analytics
+    # ------------------------------------------------------------------
+
+    def get_stats(self) -> Dict[str, Any]:
+        """Return summary statistics of the knowledge graph."""
+        G = self.graph
+
+        label_counts: Dict[str, int] = {}
+        for _, data in G.nodes(data=True):
+            lbl = data.get("label", "UNKNOWN")
+            label_counts[lbl] = label_counts.get(lbl, 0) + 1
+
+        relation_counts: Dict[str, int] = {}
+        for _, _, data in G.edges(data=True):
+            rel = data.get("relation", "UNKNOWN")
+            relation_counts[rel] = relation_counts.get(rel, 0) + 1
+
+        communities = self.get_communities()
+
+        # Degree centrality for top nodes
+        if G.number_of_nodes() > 0:
+            centrality = nx.degree_centrality(G)
+            top_nodes = sorted(centrality.items(), key=lambda x: x[1], reverse=True)[:10]
+        else:
+            top_nodes = []
+
+        return {
+            "total_nodes": G.number_of_nodes(),
+            "total_edges": G.number_of_edges(),
+            "entity_type_counts": label_counts,
+            "relation_type_counts": relation_counts,
+            "num_communities": len(communities),
+            "communities": communities,
+            "top_central_nodes": top_nodes,
+            "density": nx.density(G) if G.number_of_nodes() > 1 else 0,
+        }
+
+    def get_communities(self) -> List[List[str]]:
+        """
+        Detect communities using the greedy modularity algorithm
+        on the undirected projection.
+        """
+        if self.graph.number_of_nodes() == 0:
+            return []
+
+        undirected = self.graph.to_undirected()
+        try:
+            from networkx.algorithms.community import greedy_modularity_communities
+            communities = greedy_modularity_communities(undirected)
+            return [sorted(list(c)) for c in communities]
+        except Exception:
+            # Fallback: connected components
+            return [sorted(list(c)) for c in nx.connected_components(undirected)]
+
+    def get_node_details(self, node_id: str) -> Optional[Dict]:
+        """Return all attributes for a single node."""
+        if not self.graph.has_node(node_id):
+            return None
+        data = dict(self.graph.nodes[node_id])
+        data["id"] = node_id
+        data["in_degree"] = self.graph.in_degree(node_id)
+        data["out_degree"] = self.graph.out_degree(node_id)
+        data["neighbors"] = list(self.graph.successors(node_id)) + list(
+            self.graph.predecessors(node_id)
+        )
+        return data
+
+    # ------------------------------------------------------------------
+    # Export
+    # ------------------------------------------------------------------
+
+    def to_pyvis(self, height: str = "600px", width: str = "100%") -> Network:
+        """
+        Convert the graph to a PyVis ``Network`` for interactive
+        HTML visualisation.
+        """
+        net = Network(
+            height=height,
+            width=width,
+            directed=True,
+            bgcolor="#0a0a0a",
+            font_color="white",
+            select_menu=False,
+            filter_menu=False,
+        )
+
+        net.barnes_hut(
+            gravity=-8000,
+            central_gravity=0.3,
+            spring_length=200,
+            spring_strength=0.05,
+            damping=0.09,
+        )
+
+        for node_id, data in self.graph.nodes(data=True):
+            mentions = data.get("mentions", 1)
+            size = 15 + mentions * 5
+            net.add_node(
+                node_id,
+                label=node_id,
+                color=data.get("color", "#888888"),
+                shape=data.get("shape", "dot"),
+                size=min(size, 50),
+                title=f"{data.get('label', 'UNKNOWN')}\nMentions: {mentions}",
+                font={"size": 14, "color": "white"},
+            )
+
+        for src, tgt, data in self.graph.edges(data=True):
+            relation = data.get("relation", "")
+            weight = data.get("weight", 1)
+            net.add_edge(
+                src,
+                tgt,
+                title=relation,
+                label=relation,
+                width=min(weight * 1.5, 6),
+                color={"color": "#444444", "highlight": "#00ff88"},
+                font={"size": 10, "color": "#888888", "align": "middle"},
+                arrows={"to": {"enabled": True, "scaleFactor": 0.5}},
+                smooth={"type": "curvedCW", "roundness": 0.2},
+            )
+
+        return net
+
+    def to_dict(self) -> Dict:
+        """Serialise the graph to a JSON-safe dictionary."""
+        return nx.node_link_data(self.graph)
+
+    def from_dict(self, data: Dict) -> None:
+        """Load a graph from a dictionary produced by ``to_dict``."""
+        self.graph = nx.node_link_graph(data)

src/sample_texts.py

@@ -7,91 +7,91 @@ relationships, designed to showcase the knowledge graph pipeline.
 
 SAMPLE_TEXTS = {
 
- # ------------------------------------------------------------------
- "AI Research Paper Abstract": (
- "In recent years, the field of artificial intelligence has undergone a "
- "remarkable transformation driven by large language models. Geoffrey Hinton "
- "and Yann LeCun, both pioneers in deep learning, laid the groundwork that "
- "enabled breakthroughs such as the Transformer architecture. Their research "
- "at the University of Toronto and Meta respectively has shaped the entire "
- "landscape of modern neural networks.\n\n"
+    # ------------------------------------------------------------------
+    "AI Research Paper Abstract": (
+        "In recent years, the field of artificial intelligence has undergone a "
+        "remarkable transformation driven by large language models. Geoffrey Hinton "
+        "and Yann LeCun, both pioneers in deep learning, laid the groundwork that "
+        "enabled breakthroughs such as the Transformer architecture. Their research "
+        "at the University of Toronto and Meta respectively has shaped the entire "
+        "landscape of modern neural networks.\n\n"
 
- "OpenAI, founded in San Francisco in 2015, developed GPT-4 — a large language "
- "model that demonstrated unprecedented capabilities in natural language processing. "
- "Sam Altman, as CEO of OpenAI, oversaw the release of ChatGPT in November 2022, "
- "which reached 100 million users within two months. Meanwhile, Google DeepMind, "
- "led by Demis Hassabis in London, released AlphaFold, revolutionising protein "
- "structure prediction using deep learning.\n\n"
+        "OpenAI, founded in San Francisco in 2015, developed GPT-4 — a large language "
+        "model that demonstrated unprecedented capabilities in natural language processing. "
+        "Sam Altman, as CEO of OpenAI, oversaw the release of ChatGPT in November 2022, "
+        "which reached 100 million users within two months. Meanwhile, Google DeepMind, "
+        "led by Demis Hassabis in London, released AlphaFold, revolutionising protein "
+        "structure prediction using deep learning.\n\n"
 
- "At Stanford University in California, Fei-Fei Li and Andrew Ng pioneered "
- "computer vision and online machine learning education. Their contributions to "
- "TensorFlow and PyTorch ecosystems accelerated the adoption of neural networks "
- "across both academia and industry. The NeurIPS 2024 conference showcased over "
- "3,000 papers on topics ranging from reinforcement learning to retrieval-augmented "
- "generation.\n\n"
+        "At Stanford University in California, Fei-Fei Li and Andrew Ng pioneered "
+        "computer vision and online machine learning education. Their contributions to "
+        "TensorFlow and PyTorch ecosystems accelerated the adoption of neural networks "
+        "across both academia and industry. The NeurIPS 2024 conference showcased over "
+        "3,000 papers on topics ranging from reinforcement learning to retrieval-augmented "
+        "generation.\n\n"
 
- "Anthropic, founded by Dario Amodei in San Francisco, has focused on AI safety "
- "research. Their work on constitutional AI and RLHF (Reinforcement Learning from "
- "Human Feedback) represents a paradigm shift in how large language models are "
- "aligned with human values. The collaboration between Anthropic and Google has "
- "further strengthened the responsible AI movement in Silicon Valley."
- ),
+        "Anthropic, founded by Dario Amodei in San Francisco, has focused on AI safety "
+        "research. Their work on constitutional AI and RLHF (Reinforcement Learning from "
+        "Human Feedback) represents a paradigm shift in how large language models are "
+        "aligned with human values. The collaboration between Anthropic and Google has "
+        "further strengthened the responsible AI movement in Silicon Valley."
+    ),
 
- # ------------------------------------------------------------------
- "Tech Industry News": (
- "NVIDIA, headquartered in Santa Clara, California, reported record revenue "
- "driven by surging demand for its AI accelerator chips. Jensen Huang, CEO of "
- "NVIDIA, announced new partnerships with Microsoft and Amazon to power their "
- "cloud-based machine learning infrastructure on AWS and Azure. The company's "
- "GPU technology has become the backbone of deep learning research worldwide.\n\n"
+    # ------------------------------------------------------------------
+    "Tech Industry News": (
+        "NVIDIA, headquartered in Santa Clara, California, reported record revenue "
+        "driven by surging demand for its AI accelerator chips. Jensen Huang, CEO of "
+        "NVIDIA, announced new partnerships with Microsoft and Amazon to power their "
+        "cloud-based machine learning infrastructure on AWS and Azure. The company's "
+        "GPU technology has become the backbone of deep learning research worldwide.\n\n"
 
- "In a landmark deal, Microsoft, led by CEO Satya Nadella, invested $10 billion "
- "in OpenAI in January 2023, securing exclusive rights to integrate GPT-4 into "
- "its products. GitHub Copilot, powered by OpenAI's Codex model, has transformed "
- "software development, with over 1 million developers using the AI-powered "
- "coding assistant built on the Transformer architecture.\n\n"
+        "In a landmark deal, Microsoft, led by CEO Satya Nadella, invested $10 billion "
+        "in OpenAI in January 2023, securing exclusive rights to integrate GPT-4 into "
+        "its products. GitHub Copilot, powered by OpenAI's Codex model, has transformed "
+        "software development, with over 1 million developers using the AI-powered "
+        "coding assistant built on the Transformer architecture.\n\n"
 
- "Apple, under Tim Cook, has taken a different approach to artificial intelligence, "
- "focusing on on-device machine learning and privacy-first AI. At their headquarters "
- "in Cupertino, California, Apple's research team has developed advanced natural "
- "language processing models that run entirely on iPhone hardware without cloud "
- "connectivity. Samsung and Qualcomm have pursued similar edge AI strategies.\n\n"
+        "Apple, under Tim Cook, has taken a different approach to artificial intelligence, "
+        "focusing on on-device machine learning and privacy-first AI. At their headquarters "
+        "in Cupertino, California, Apple's research team has developed advanced natural "
+        "language processing models that run entirely on iPhone hardware without cloud "
+        "connectivity. Samsung and Qualcomm have pursued similar edge AI strategies.\n\n"
 
- "The European Union introduced the AI Act in 2024, establishing the world's first "
- "comprehensive regulatory framework for artificial intelligence. The legislation "
- "affects major tech companies operating in Europe, including Google, Meta, and "
- "Amazon. Mark Zuckerberg responded by open-sourcing Meta's LLaMA models, while "
- "Sundar Pichai pledged Google's commitment to responsible AI development in "
- "partnership with UNESCO and the United Nations."
- ),
+        "The European Union introduced the AI Act in 2024, establishing the world's first "
+        "comprehensive regulatory framework for artificial intelligence. The legislation "
+        "affects major tech companies operating in Europe, including Google, Meta, and "
+        "Amazon. Mark Zuckerberg responded by open-sourcing Meta's LLaMA models, while "
+        "Sundar Pichai pledged Google's commitment to responsible AI development in "
+        "partnership with UNESCO and the United Nations."
+    ),
 
- # ------------------------------------------------------------------
- "Historical Event": (
- "The development of computing in the 20th century stands as one of humanity's "
- "greatest intellectual achievements. Alan Turing, working at Cambridge University "
- "in the 1930s, formulated the concept of a universal computing machine that could "
- "simulate any algorithmic process. His work at Bletchley Park during World War II "
- "on breaking the Enigma cipher demonstrated the practical power of computation.\n\n"
+    # ------------------------------------------------------------------
+    "Historical Event": (
+        "The development of computing in the 20th century stands as one of humanity's "
+        "greatest intellectual achievements. Alan Turing, working at Cambridge University "
+        "in the 1930s, formulated the concept of a universal computing machine that could "
+        "simulate any algorithmic process. His work at Bletchley Park during World War II "
+        "on breaking the Enigma cipher demonstrated the practical power of computation.\n\n"
 
- "John von Neumann, a mathematician at the Institute for Advanced Study in "
- "Princeton, designed the architecture that bears his name in 1945. The von Neumann "
- "architecture — separating memory, processing, and input/output — remains the "
- "foundation of virtually every modern computer. Claude Shannon, working at Bell "
- "Labs in New Jersey, simultaneously developed information theory, establishing "
- "the mathematical framework for digital communication.\n\n"
+        "John von Neumann, a mathematician at the Institute for Advanced Study in "
+        "Princeton, designed the architecture that bears his name in 1945. The von Neumann "
+        "architecture — separating memory, processing, and input/output — remains the "
+        "foundation of virtually every modern computer. Claude Shannon, working at Bell "
+        "Labs in New Jersey, simultaneously developed information theory, establishing "
+        "the mathematical framework for digital communication.\n\n"
 
- "Ada Lovelace, often regarded as the first computer programmer, wrote algorithms "
- "for Charles Babbage's Analytical Engine in London during the 1840s. Her vision "
- "of machines that could manipulate symbols beyond mere calculation anticipated "
- "modern artificial intelligence by over a century. The United Kingdom has since "
- "honored her legacy with Ada Lovelace Day, celebrated annually in October.\n\n"
+        "Ada Lovelace, often regarded as the first computer programmer, wrote algorithms "
+        "for Charles Babbage's Analytical Engine in London during the 1840s. Her vision "
+        "of machines that could manipulate symbols beyond mere calculation anticipated "
+        "modern artificial intelligence by over a century. The United Kingdom has since "
+        "honored her legacy with Ada Lovelace Day, celebrated annually in October.\n\n"
 
- "The founding of the Internet can be traced to ARPANET, developed by the United "
- "States Department of Defense in the 1960s. Tim Berners-Lee, working at CERN in "
- "Switzerland, invented the World Wide Web in 1989. His creation transformed how "
- "humanity shares knowledge. Today, the convergence of quantum computing, "
- "artificial intelligence, and blockchain technology promises yet another "
- "revolution, with IBM, Google, and Microsoft leading the charge in quantum "
- "computing research from laboratories across North America and Europe."
- ),
+        "The founding of the Internet can be traced to ARPANET, developed by the United "
+        "States Department of Defense in the 1960s. Tim Berners-Lee, working at CERN in "
+        "Switzerland, invented the World Wide Web in 1989. His creation transformed how "
+        "humanity shares knowledge. Today, the convergence of quantum computing, "
+        "artificial intelligence, and blockchain technology promises yet another "
+        "revolution, with IBM, Google, and Microsoft leading the charge in quantum "
+        "computing research from laboratories across North America and Europe."
+    ),
 }

src/visualizer.py

@@ -18,28 +18,28 @@ from src.graph_builder import KnowledgeGraph, ENTITY_COLORS
 # ------------------------------------------------------------------
 
 def create_pyvis_graph(kg: KnowledgeGraph, height: str = "650px") -> str:
- """
- Render *kg* as an interactive PyVis graph and return raw HTML.
-
- The HTML string can be embedded directly with
- ``streamlit.components.v1.html()``.
- """
- net = kg.to_pyvis(height=height)
-
- # Generate HTML string (PyVis >= 0.3 supports generate_html)
- try:
- html = net.generate_html()
- except AttributeError:
- # Fallback for older pyvis
- import tempfile, os
- tmp = tempfile.NamedTemporaryFile(delete=False, suffix=".html", mode="w", encoding="utf-8")
- net.save_graph(tmp.name)
- tmp.close()
- with open(tmp.name, "r", encoding="utf-8") as f:
- html = f.read()
- os.unlink(tmp.name)
-
- return html
+    """
+    Render *kg* as an interactive PyVis graph and return raw HTML.
+
+    The HTML string can be embedded directly with
+    ``streamlit.components.v1.html()``.
+    """
+    net = kg.to_pyvis(height=height)
+
+    # Generate HTML string (PyVis >= 0.3 supports generate_html)
+    try:
+        html = net.generate_html()
+    except AttributeError:
+        # Fallback for older pyvis
+        import tempfile, os
+        tmp = tempfile.NamedTemporaryFile(delete=False, suffix=".html", mode="w", encoding="utf-8")
+        net.save_graph(tmp.name)
+        tmp.close()
+        with open(tmp.name, "r", encoding="utf-8") as f:
+            html = f.read()
+        os.unlink(tmp.name)
+
+    return html
 
 
 # ------------------------------------------------------------------
@@ -47,146 +47,146 @@ def create_pyvis_graph(kg: KnowledgeGraph, height: str = "650px") -> str:
 # ------------------------------------------------------------------
 
 _CHART_LAYOUT = dict(
- paper_bgcolor="#0a0a0a",
- plot_bgcolor="#111111",
- font_color="white",
- font_size=13,
- margin=dict(l=40, r=40, t=50, b=40),
+    paper_bgcolor="#0a0a0a",
+    plot_bgcolor="#111111",
+    font_color="white",
+    font_size=13,
+    margin=dict(l=40, r=40, t=50, b=40),
 )
 
 
 def graph_stats_chart(stats: Dict[str, Any]) -> go.Figure:
- """
- Create a combined Plotly figure with:
- - Entity type distribution (bar)
- - Relationship type distribution (bar)
- """
- entity_counts = stats.get("entity_type_counts", {})
- relation_counts = stats.get("relation_type_counts", {})
-
- fig = make_subplots(
- rows=1,
- cols=2,
- subplot_titles=("Entity Types", "Relationship Types"),
- horizontal_spacing=0.15,
- )
-
- # --- Entity type bar chart ---
- if entity_counts:
- types = list(entity_counts.keys())
- counts = list(entity_counts.values())
- colors = [ENTITY_COLORS.get(t, "#888888") for t in types]
-
- fig.add_trace(
- go.Bar(
- x=types,
- y=counts,
- marker_color=colors,
- text=counts,
- textposition="outside",
- name="Entity Types",
- showlegend=False,
- ),
- row=1,
- col=1,
- )
-
- # --- Relationship type bar chart ---
- if relation_counts:
- rels = list(relation_counts.keys())
- rcounts = list(relation_counts.values())
-
- fig.add_trace(
- go.Bar(
- x=rels,
- y=rcounts,
- marker_color="#00d4ff",
- text=rcounts,
- textposition="outside",
- name="Relationships",
- showlegend=False,
- ),
- row=1,
- col=2,
- )
-
- fig.update_layout(
- height=370,
- **_CHART_LAYOUT,
- )
- fig.update_xaxes(tickangle=-40)
-
- return fig
+    """
+    Create a combined Plotly figure with:
+      - Entity type distribution (bar)
+      - Relationship type distribution (bar)
+    """
+    entity_counts = stats.get("entity_type_counts", {})
+    relation_counts = stats.get("relation_type_counts", {})
+
+    fig = make_subplots(
+        rows=1,
+        cols=2,
+        subplot_titles=("Entity Types", "Relationship Types"),
+        horizontal_spacing=0.15,
+    )
+
+    # --- Entity type bar chart ---
+    if entity_counts:
+        types = list(entity_counts.keys())
+        counts = list(entity_counts.values())
+        colors = [ENTITY_COLORS.get(t, "#888888") for t in types]
+
+        fig.add_trace(
+            go.Bar(
+                x=types,
+                y=counts,
+                marker_color=colors,
+                text=counts,
+                textposition="outside",
+                name="Entity Types",
+                showlegend=False,
+            ),
+            row=1,
+            col=1,
+        )
+
+    # --- Relationship type bar chart ---
+    if relation_counts:
+        rels = list(relation_counts.keys())
+        rcounts = list(relation_counts.values())
+
+        fig.add_trace(
+            go.Bar(
+                x=rels,
+                y=rcounts,
+                marker_color="#00d4ff",
+                text=rcounts,
+                textposition="outside",
+                name="Relationships",
+                showlegend=False,
+            ),
+            row=1,
+            col=2,
+        )
+
+    fig.update_layout(
+        height=370,
+        **_CHART_LAYOUT,
+    )
+    fig.update_xaxes(tickangle=-40)
+
+    return fig
 
 
 def centrality_chart(top_nodes: list) -> go.Figure:
- """
- Horizontal bar chart of the top-N most central nodes.
- """
- if not top_nodes:
- fig = go.Figure()
- fig.update_layout(
- title="No nodes to display",
- **_CHART_LAYOUT,
- height=300,
- )
- return fig
-
- names = [n[0] for n in reversed(top_nodes)]
- values = [round(n[1], 4) for n in reversed(top_nodes)]
-
- fig = go.Figure(
- go.Bar(
- x=values,
- y=names,
- orientation="h",
- marker=dict(
- color=values,
- colorscale=[[0, "#0a0a0a"], [0.5, "#00d4ff"], [1, "#00ff88"]],
- ),
- text=[f"{v:.3f}" for v in values],
- textposition="outside",
- )
- )
-
- fig.update_layout(
- title="Top Nodes by Degree Centrality",
- xaxis_title="Centrality Score",
- height=max(300, len(top_nodes) * 35 + 100),
- **_CHART_LAYOUT,
- )
-
- return fig
+    """
+    Horizontal bar chart of the top-N most central nodes.
+    """
+    if not top_nodes:
+        fig = go.Figure()
+        fig.update_layout(
+            title="No nodes to display",
+            **_CHART_LAYOUT,
+            height=300,
+        )
+        return fig
+
+    names = [n[0] for n in reversed(top_nodes)]
+    values = [round(n[1], 4) for n in reversed(top_nodes)]
+
+    fig = go.Figure(
+        go.Bar(
+            x=values,
+            y=names,
+            orientation="h",
+            marker=dict(
+                color=values,
+                colorscale=[[0, "#0a0a0a"], [0.5, "#00d4ff"], [1, "#00ff88"]],
+            ),
+            text=[f"{v:.3f}" for v in values],
+            textposition="outside",
+        )
+    )
+
+    fig.update_layout(
+        title="Top Nodes by Degree Centrality",
+        xaxis_title="Centrality Score",
+        height=max(300, len(top_nodes) * 35 + 100),
+        **_CHART_LAYOUT,
+    )
+
+    return fig
 
 
 def community_chart(communities: list) -> go.Figure:
- """Pie chart showing community sizes."""
- if not communities:
- fig = go.Figure()
- fig.update_layout(title="No communities detected", **_CHART_LAYOUT, height=300)
- return fig
-
- labels = [f"Community {i+1}" for i in range(len(communities))]
- sizes = [len(c) for c in communities]
-
- fig = go.Figure(
- go.Pie(
- labels=labels,
- values=sizes,
- hole=0.45,
- marker=dict(
- colors=["#00ff88", "#00d4ff", "#a855f7", "#f59e0b", "#ec4899",
- "#6366f1", "#14b8a6", "#f43f5e", "#84cc16", "#06b6d4"],
- ),
- textinfo="label+percent",
- textfont_size=12,
- )
- )
-
- fig.update_layout(
- title="Community Distribution",
- height=370,
- **_CHART_LAYOUT,
- )
-
- return fig
+    """Pie chart showing community sizes."""
+    if not communities:
+        fig = go.Figure()
+        fig.update_layout(title="No communities detected", **_CHART_LAYOUT, height=300)
+        return fig
+
+    labels = [f"Community {i+1}" for i in range(len(communities))]
+    sizes = [len(c) for c in communities]
+
+    fig = go.Figure(
+        go.Pie(
+            labels=labels,
+            values=sizes,
+            hole=0.45,
+            marker=dict(
+                colors=["#00ff88", "#00d4ff", "#a855f7", "#f59e0b", "#ec4899",
+                         "#6366f1", "#14b8a6", "#f43f5e", "#84cc16", "#06b6d4"],
+            ),
+            textinfo="label+percent",
+            textfont_size=12,
+        )
+    )
+
+    fig.update_layout(
+        title="Community Distribution",
+        height=370,
+        **_CHART_LAYOUT,
+    )
+
+    return fig