update metadata

app/app.py

@@ -13,7 +13,7 @@ with open('/app/data/redis_data.json', 'r') as file:
 
 skill_details_mapping = {}
 
-df_course = pd.read_csv('/app/data/all_course_info.csv')
+
 # Function to retrieve documents based on selected skills
 def retrieve_documents(occupation,skills):
     output = []
@@ -36,7 +36,7 @@ def retrieve_documents(occupation,skills):
     print(query)
     docs = retriever.get_relevant_documents(query)
 
-    partial_compare_docs = functools.partial(compare_docs_with_context, df_course=df_course, target_occupation_name=target_occupation_name, target_occupation_dsp=target_occupation_dsp,skill_gap = skill_query)
+    partial_compare_docs = functools.partial(compare_docs_with_context, target_occupation_name=target_occupation_name, target_occupation_dsp=target_occupation_dsp,skill_gap = skill_query)
     sorted_docs = sorted(docs, key=functools.cmp_to_key(partial_compare_docs), reverse=True)
 
     

app/data_process.py

@@ -35,11 +35,13 @@ def build_occupation_query(occupation):
     occupation_name_de = occupation['preferredLabel'].get('de','')
     occupation_dsp = occupation['description'].get('de','').get('literal','')
     occupation_query = occupation_name_de +" " + occupation['preferredLabel'].get('en','')+" "+ occupation['description'].get('de','').get('literal','') + " "+ occupation_dsp
+    '''
     if occupation['_links']['broaderIscoGroup']:
         for group in occupation['_links']['broaderIscoGroup']:
             occupation_query += " " + group['title']
     else:
         pass
+        '''
     return occupation_name_de,occupation_dsp,occupation_query      
 
 # Get occupations from a CSV

app/embedding_setup.py

@@ -126,15 +126,9 @@ def evaluate(
         output = [_.split('Response:\n')[-1] for _ in output]
         return output, logits.tolist()
 
-def compare_docs_with_context(doc_a, doc_b, df_course, target_occupation_name, target_occupation_dsp,skill_gap):
-    # Extract course details from the data frame
-    course_a = df_course[df_course['course_id'] == int(doc_a.metadata['id'])].iloc[0]
-    course_b = df_course[df_course['course_id'] == int(doc_b.metadata['id'])].iloc[0]
-    print('comapring...')
-    print(course_a['course_name'], course_b['course_name'])
-    # Prepare the input for chain_re.invoke
+def compare_docs_with_context(doc_a, doc_b, target_occupation_name, target_occupation_dsp,skill_gap):
     
-    courses = f"First: name: {course_a['course_name']}  description:{course_a['course_content_limited']} Second: name: {course_b['course_name']}  description:{course_b['course_content_limited']}" 
+    courses = f"First: name: {doc_a.metadata['name']}  description:{doc_a.metadata['description']} Second: name: {doc_b.metadata['name']}  description:{doc_b.metadata['description']}" 
     #courses = f"First: name: {course_a['course_name']}  skills:{course_a['course_skills_edu']} Second: name: {course_b['course_name']}  skills:{course_b['course_skills_edu']}" 
     target_occupation = f"name: {target_occupation_name} description: {target_occupation_dsp}"
     skill_gap = skill_gap