Update README.md

README.md

@@ -1,13 +1,21 @@
----
-title: Country Specific Updated Sentiment Analysis Of India
-emoji: 🚀
-colorFrom: purple
-colorTo: red
-sdk: streamlit
-sdk_version: 1.42.2
-app_file: app.py
-pinned: false
-short_description: Analyzing 5,000+ MEA India press releases using semantic mat
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Country-specific-Updated-Sentiment-Analysis-of-India
+ Analyzing 5,000+ MEA India press releases using semantic matching via embeddings
+```mermaid
+flowchart TD
+    A[Start] --> B[Load CSV Data: Read mea_press_releases_1.csv]
+    B --> C[Preprocess Data:<br>- Convert date to datetime<br>- Combine title and description into combined_text<br>- Drop rows with missing values<br>- Extract year from date]
+    C --> D[Calculate Sentiment Scores:<br>- Initialize VADER Sentiment Analyzer<br>- Apply to combined_text for all rows]
+    D --> E[Extract Countries:<br>- Use spaCy NER to identify countries in combined_text<br>- Remove 'India' and 'New Delhi'<br>- Store as comma-separated strings in countries column]
+    E --> F[Get Unique Country Names from countries column]
+    F --> G[Generate Embeddings:<br>- Use SentenceTransformer to encode unique countries]
+    G --> H[Build FAISS Index with country embeddings]
+    H --> I[Input Query:<br>- Country name e.g., 'United States'<br>- Year threshold e.g., 2010]
+    I --> J[Encode Query Using SentenceTransformer]
+    J --> K[Find Similar Country Names:<br>- Search FAISS index for top k matches]
+    K --> L[Filter Data:<br>- Match countries column with similar names<br>- Keep rows where year >= threshold]
+    L --> M[Calculate Mean Sentiment per Year]
+    M --> N[Plot Mean Sentiment Over Years:<br>- Line plot with yearly averages]
+    N --> O[Plot Sentiment Over Time:<br>- Scatter plot with color-coded sentiments]
+    O --> P[Calculate and Print Average Sentiment]
+    P --> Q[End]
+   ```