Update README.md

README.md

@@ -18,11 +18,11 @@ The parameters passed in the url request are :
 - fq=publicationDateY_i:[2013%20TO%202023]
 - fl=halId_s,doiId_s,uri_s,title_s,subTitle_s,authFullName_s,producedDate_s,journalTitle_s,journalPublisher_s,abstract_s,fr_keyword_s,openAccess_bool,submitType_s
 
+The combined column conatins a concatenation of the textual contents of the three columns : title_s, subTitle_s and abstract_s.
 The embeddings corpus hal_embeddings.pkl stores the embeddings of the "combined" column values converted in vectors with the sentence-transformers/all-MiniLM-L6-v2 embeddings model.
 
 Furthermore, all the dataset (except the "abstract" and "combined" columns) has been converted in a Knowledge Graph and stored in a Neo4j Graph store which persists texts and embeddings.
 The text embeddings model used is nomic-embed-text-v1.5.
-The Knowledge Graph Index is persiste
 
 
 ## Metadata extraction
@@ -157,3 +157,67 @@ article_data_list
 
 The KnowledgeGraphIndex is persisted in the /index_storage folder, and can be easely reloaded in a Neo4j database and/or reloaded to be queried by a LlamaIndex KnowledgeGraphQueryEngine.
 
+```
+import pandas as pd
+from datasets import load_dataset
+from llama_index.core import SimpleDirectoryReader, KnowledgeGraphIndex, StorageContext, load_index_from_storage
+from llama_index.graph_stores.neo4j import Neo4jGraphStore
+from llama_index.vector_stores.neo4jvector import Neo4jVectorStore
+from llama_index.embeddings.nomic import NomicEmbedding
+from llama_index.llms.groq import Groq
+from llama_index.core import Settings
+import nest_asyncio
+
+# Load the dataset
+hal_data = load_dataset("Geraldine/hal_univcotedazur_shs_articles_2013-2023", data_files="hal_data.csv")
+df = pd.DataFrame(hal_data["train"])
+df = df.drop(columns=["abstract_s","combined"])
+df.to_csv("hal_data.csv", index=False, encoding="utf-8")
+
+# Document reader
+reader = SimpleDirectoryReader(input_files=["./hal_data.csv"])
+documents = reader.load_data()
+
+# Embeddings & LLM
+NOMIC_API_KEY = "..."
+GROQ_API_KEY = "..."
+
+nest_asyncio.apply()
+
+embed_model = NomicEmbedding(
+    api_key=NOMIC_API_KEY,
+    dimensionality=768,
+    model_name="nomic-embed-text-v1.5",
+)
+
+llm = Groq(model="mixtral-8x7b-32768", api_key=GROQ_API_KEY)
+
+Settings.llm = llm
+Settings.embed_model = embed_model
+Settings.chunk_size = 512
+
+# Neo4j Graph store & KnowledgeGraph index creation
+graph_store = Neo4jGraphStore(
+    username="...",
+    password="...",
+    url="...",
+)
+
+storage_context = StorageContext.from_defaults(graph_store=graph_store)
+
+index = KnowledgeGraphIndex.from_documents(
+    documents,
+    storage_context=storage_context,
+    include_embeddings=True,
+    max_triplets_per_chunk=2,
+)
+
+# Persist index
+index.storage_context.persist("./index_storage")
+
+# Reload index
+storage_context = StorageContext.from_defaults(persist_dir="./index_storage")
+index = load_index_from_storage(storage_context)
+```
+
+