#!/usr/bin/env python
# coding: utf-8

# # AI Research Summarizer with LLM Feedback
# 
# Overview:
# 
# The AI Research Summarizer is an innovative tool designed to help researchers, students, and AI enthusiasts stay up-to-date with the latest developments in artificial intelligence research. By leveraging the power of the arXiv API and a local language model (LLM), specifically the Llama 3 model, this application fetches the most recent AI research papers and provides concise summaries along with insightful feedback from the LLM.
# 
# Key Features:
# 
# - Fetches the latest AI research papers from arXiv using the API
# - Generates concise summaries of the research paper abstracts
# - Provides LLM-generated thoughts and feedback on the research topics
# - Presents the summaries, thoughts, and metadata in a readable format
# - Offers customization options for search queries and the number of results
# 
# Benefits:
# 
# - Time-saving: The AI Research Summarizer saves researchers and AI enthusiasts valuable time by automatically fetching and summarizing the latest research papers, eliminating the need to manually search and read through numerous abstracts.
# - Insightful feedback: The unique feature of this application is the incorporation of LLM-generated thoughts and feedback on the research topics. The Llama 3 model, trained on a vast corpus of text, provides valuable insights and perspectives on the summarized research. This feedback helps users quickly grasp the significance and potential implications of the research, making the tool more robust and informative.
# - Personalized experience: Users can customize their search queries and the number of results to focus on specific areas of interest within AI research. This personalization enhances the user experience and ensures that the summaries and feedback are tailored to their needs.
# - Accessibility: The AI Research Summarizer makes AI research more accessible to a wider audience. By providing concise summaries and LLM-generated thoughts, the tool helps bridge the gap between highly technical research papers and a broader audience interested in understanding the latest developments in AI.
# - Inspiration and ideation: The LLM-generated thoughts and feedback can inspire new ideas and research directions. By offering a different perspective on the research topics, the application encourages users to think critically and explore novel approaches to AI challenges.
# - Efficiency: The application leverages the power of APIs and local LLMs to efficiently fetch, summarize, and analyze research papers. This efficiency allows users to quickly stay informed about the latest AI research without the need for extensive computational resources.
# 
# In conclusion, the AI Research Summarizer with LLM Feedback is a powerful tool that combines the capabilities of the arXiv API and the Llama 3 model to provide users with concise summaries and valuable insights into the latest AI research. By saving time, offering personalized experiences, and inspiring new ideas, this application is a valuable asset for researchers, students, and AI enthusiasts alike. The incorporation of LLM-generated feedback sets this tool apart, making it a robust and informative resource for anyone interested in staying at the forefront of AI research.

# In[21]:


from langchain.llms import Ollama
import requests
import xml.etree.ElementTree as ET

def summarize_text(text, llm):
    prompt = f"Provide a summary of the following research paper abstract and your thoughts on the subject:\n\n{text}"
    summary = llm(prompt)
    return summary.strip()

def search_and_summarize_ai_research():
    # URL of the arXiv API for searching AI research papers, sorted by submission date in descending order
    url = "https://export.arxiv.org/api/query?search_query=all:ai&start=0&max_results=5&sortBy=submittedDate&sortOrder=descending"

    # Send a GET request to the API URL and parse the XML response
    response = requests.get(url)
    root = ET.fromstring(response.content)

    # Initialize the Ollama LLM
    llm = Ollama(model="llama3:latest", stop=["<|eot_id|>"])

    # Extract the summaries and metadata from the XML entries
    summaries = []
    for entry in root.findall("{http://www.w3.org/2005/Atom}entry"):
        title = entry.find("{http://www.w3.org/2005/Atom}title").text
        summary = entry.find("{http://www.w3.org/2005/Atom}summary").text
        link = entry.find("{http://www.w3.org/2005/Atom}link[@type='text/html']").attrib["href"]
        published = entry.find("{http://www.w3.org/2005/Atom}published").text

        # Generate a concise summary of the paper using the Ollama LLM
        concise_summary = summarize_text(summary, llm)

        summaries.append({
            "title": title,
            "summary": concise_summary,
            "link": link,
            "published": published
        })

    return summaries

# Print the summaries of the latest AI research papers
summaries = search_and_summarize_ai_research()
for summary in summaries:
    print(f"Title: {summary['title']}")
    print(f"Published: {summary['published']}")
    print(f"Summary: {summary['summary']}")
    print(f"\nLink: {summary['link']}")
    print("\n---\n")