added all the remaining categories

app.py

@@ -117,38 +117,59 @@ def get_response_from_chatboat(message,chat_history, max_convo_length=10):
 
 
 def analyse_current_conversation(text, analysis_type):
-    
-    if(ANALYSIS_TYPES.get(analysis_type, None) is None):
-        return f"Analysis type {analysis_type} is not implemented yet, please choose another category"
-    
-    if not text:
-        return f"No text provided to analyze for {analysis_type}, please provide text or load from chatboat history"
-    
-    word_count = len(text.split())
 
-    if(word_count < 20 ):
-        return f" The text is too short to analyze for {analysis_type}, please provide a large text"
+    try:
+        if(ANALYSIS_TYPES.get(analysis_type, None) is None):
+            return f"Analysis type {analysis_type} is not implemented yet, please choose another category"
+        
+        if not text:
+            return f"No text provided to analyze for {analysis_type}, please provide text or load from chatboat history"
+        
+        word_count = len(text.split())
+
+        if(word_count < 20 ):
+            return f" The text is too short to analyze for {analysis_type}, please provide a large text"
+
+        system_prompt = get_system_prompt(analysis_type)
+        text_to_analyze = text
+
+        response = client.chat.completions.create(
+        messages=[
+            {"role": "system", "content": system_prompt},
+            {"role": "user", "content": text_to_analyze}
+        ],
+        model="gpt-3.5-turbo",
+        )
+        
+        analysis_result = response.choices[0].message.content  
+        print(analysis_result)
+
+        # Parse the result, handle JSON parsing errors
+        try:
+            parsed_result = json.loads(analysis_result)
+        except json.JSONDecodeError:
+            return "Failed to parse the analysis result. Please check the format of the returned data."
+
+        formatted_json = format_result_to_markdown(parsed_result)
+        return formatted_json
+
+    except KeyError as e:
+        return f"Key error occurred: {e}. Please check your keys."
+    except Exception as e:
+        # Check if the error message is related to the API key
+        if 'API key' in str(e):
+            return "OpenAI API key error: Please verify your API key."
+        else:
+            return f"An unexpected error occurred: {e}. Please check your implementation."
 
-    system_prompt = get_system_prompt(analysis_type)
-    text_to_analyze = text
 
-    response = client.chat.completions.create(
-    messages=[
-        {"role": "system", "content": system_prompt},
-        {"role": "user", "content": text_to_analyze}
-    ],
-    model="gpt-3.5-turbo",
-    )
-    
-    analysis_result = response.choices[0].message.content  
-    print(analysis_result)
-    parsed_result = json.loads(analysis_result)
+    # parsed_result = json.loads(analysis_result)
 
-    formated_json = format_result_to_markdown(parsed_result)
+    # formated_json = format_result_to_markdown(parsed_result)
 
-    print(parsed_result)
-    # Your implementation for counting words and performing analysis
-    return formated_json
+    # print(parsed_result)
+    # # Your implementation for counting words and performing analysis
+    # return formated_json
 
 
 

utils.py

@@ -12,9 +12,9 @@ ai_audit_analysis_categories = {
         ],
 
     "GDPR": [
-        "Consent_and_Transparency", 
-        "Data_Security", 
         "Privacy_Assessment",
+        "Consent_and_Transparency", 
+        "Data_Security",
         "Environmental_Impact"],
 
     "Toxicity": [
@@ -61,10 +61,6 @@ ai_audit_analysis_categories = {
 }
 
 
-
-
-
-
 # Define a standard template for prompts
 STANDARD_PROMPT_TEMPLATE = "You are a data analysis assistant capable of {analysis_type} analysis. {specific_instruction} Respond with your analysis in JSON format. The JSON schema should include '{json_schema}'."
 
@@ -90,11 +86,59 @@ ANALYSIS_TYPES = {
     "stress_level_detection": "Analyze the text to assess stress levels, identifying triggers and intensity of stress.",
     "empathy_level_assessment": "Assess the level of empathy expressed in the text, identifying empathetic responses and tendencies.",
     "mood_detection": "Detect the mood of the individual based on textual cues, ranging from happy to sad, calm to angry.",
-    "toxicity_detection": "Identify and assess the level of toxicity in the provided text. Determine whether the text contains harmful, offensive, or inappropriate content and provide a score indicating the severity of the toxicity"
+    "toxicity_detection": "Identify and assess the level of toxicity in the provided text. Determine whether the text contains harmful, offensive, or inappropriate content and provide a score indicating the severity of the toxicity",
+
+     # GDPR-related types
+    "Consent_and_Transparency": "Evaluate how consent is obtained and the level of transparency provided to users regarding data usage.",
+    "Data_Security": "Assess the measures in place for data security, including vulnerabilities and compliance with security standards.",
+    "Privacy_Assessment": "Analyze the overall privacy practices, including policy compliance, data minimization, and user data accessibility.",
+    "Environmental_Impact": "Assess the environmental impact of data processing practices, including carbon footprint and energy efficiency.",
+
+    # Toxicity-related types
+    "Content_Moderation": "Evaluate the effectiveness of content moderation practices, including automated and human moderation efforts.",
+    "Reporting_Mechanism": "Assess the ease and effectiveness of reporting mechanisms for inappropriate or harmful content.",
+    "Content_Guidelines": "Analyze the clarity and comprehensiveness of content guidelines and their enforcement consistency.",
+    "User_Education": "Evaluate the availability and accessibility of educational resources for users regarding appropriate content and behavior.",
+
+    # Legal-related types
+    "Privacy_Policy": "Analyze the clarity and compliance of a privacy policy with legal standards.",
+    "Data_Retention": "Evaluate the data retention practices, including periods, deletion policies, and legal compliance.",
+    "Consent_Mechanism": "Assess the clarity and effectiveness of the consent mechanism in place for data collection and usage.",
+    "GDPR_Compliance": "Evaluate the level of GDPR compliance in data handling, protection measures, and breach notification protocols.",
+
+    # Context-related types
+    "Ethical_AI": "Assess adherence to ethical standards in AI practices, including identification and mitigation of ethical issues.",
+    "Bias_Mitigation": "Evaluate the presence and mitigation of bias in data or algorithms.",
+    "Fairness_Assessment": "Assess fairness in AI systems, identifying affected groups and providing recommendations for improvement.",
+    "Explainability": "Evaluate the transparency and explainability of AI models to users.",
+
+    # Governance-related types
+    "Model_Development": "Analyze the process of model development, including team composition and ethical considerations.",
+    "Data_Quality": "Assess the quality of data used, focusing on accuracy, completeness, and timeliness.",
+    "User_Input": "Evaluate the mechanisms for and impact of user feedback on the system.",
+
+    # Risk Management-related types
+    "Corporate_Ethics": "Assess the ethical practices within a corporation, including employee training and ethics code adherence.",
+    "Board_Management": "Evaluate the effectiveness and diversity of board management and its compliance with ethical standards.",
+    "Stakeholder_Engagement": "Analyze stakeholder engagement practices, including inclusion, feedback mechanisms, and satisfaction.",
+    "Risk_Management": "Assess the identification, mitigation, and monitoring of risks within an organization.",
+
+    # Robustness-related types
+    "System_Reliability": "Evaluate the reliability and resilience of a system, including uptime and redundancy measures.",
+    "Quality_Assurance": "Assess the quality assurance practices, including compliance with standards and testing frequency.",
+    "Stress_Testing": "Analyze the system's robustness through stress testing and identify weaknesses.",
+    "Fail_Safe_Procedures": "Evaluate the effectiveness of fail-safe procedures in place for system failures.",
+
+    # Sustainability-related types
+    "Renewable_Resources": "Assess the use of renewable resources and sustainability goals in operations.",
+    "Waste_Reduction": "Evaluate waste management practices, reduction rates, and recycling initiatives.",
+    "Energy_Efficiency": "Analyze energy consumption and efficiency, including energy-saving measures and audits.",
+    "Sustainable_Practices": "Evaluate the adoption of sustainable practices, including training and overall impact."
 }
 
 
 JSON_SCHEMAS = {
+
     "sentiment_analysis": {
          "sentiment": "string (positive, negative, neutral)",
          "confidence_score": "number (0-1)",
@@ -304,4 +348,3 @@ JSON_SCHEMAS = {
 
 
 
-