Update README

README2.md

+# To build the Docker image (run this only once)
+sudo docker build -t source-extractor .
+
+# To run the source-extractor, you can take "sample_input" as <input_folder>
+(sudo) ./run_docker.sh extract <input_folder>
+
+# To evaluate the model's performance, the folder input_labeled contains 363 annotated documents and is provided upon request
+Install python3 
+pip3 install sklearn
+(sudo) ./run_docker.sh evaluate input_labeled
+
+Performance report:
+SOURCE-PRIM
+F1 score 0.971039182282794
+Recall 0.9595959595959596
+Precision 0.9827586206896551
+SOURCE-SEC
+F1 score 0.8920863309352518
+Recall 0.8378378378378378
+Precision 0.9538461538461539