Tool

Search and Load Open Access Papers from OpenAlex

Search OpenAlex for papers, including full-text search of Open Access works, into LlamaIndex.

Works with openalex

73
Spark score
out of 100
Updated 2 days ago
Version 0.14.23
Models

Add to Favorites

Why it matters

Access and process scholarly literature from OpenAlex, enabling full-text searches for open access papers to support research and analysis.

Outcomes

What it gets done

01

Search OpenAlex for academic papers based on a query.

02

Load paper data into a structured format for further processing.

03

Enable full-text search capabilities for open access publications.

04

Retrieve answers with citations from the loaded paper data.

Install

Add it to your toolbox

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/li-reader-readers-openalex | bash

Overview

OpenAlex Reader

A LlamaIndex reader that searches OpenAlex for papers, with optional full-text search of Open Access works. Use for citation-backed research queries, and full_text=True specifically for searching inside Open Access paper content.

What it does

The OpenAlex Reader searches for papers in OpenAlex and loads them into LlamaIndex. Its main advantage over similar loaders is that it can search the full text of Open Access papers, not just titles and abstracts, when that option is enabled.

OpenAlexReader is initialized with an email address, and load_data takes a query string plus a full_text boolean - set to True to search full-text content of Open Access papers, or left False for the default metadata-level search. The source's own worked example queries "biases in large language models" and, once indexed, a query engine call produces a markdown table of biases with citations - showing results like data selection bias, social bias across several categories, and anti-Muslim and anti-LGBTQ+ bias, each attributed to a specific numbered source.

When to use - and when NOT to

Use it when you need to search and retrieve academic papers - grounded with citations to specific sources - for a research topic, and specifically use full_text=True when abstract-level metadata isn't enough and you need to search inside the actual paper content of Open Access works. Do not expect full-text search to work for papers that are not Open Access; that capability is scoped to openly accessible content.

Capabilities

load_data searches OpenAlex for papers matching a query, optionally searching full text for Open Access papers, and returns them as LlamaIndex documents ready for citation-backed querying.

How to install

pip install llama-index-readers-openalex

Requires an email address passed to OpenAlexReader (OpenAlex's API convention for identifying callers).

Who it's for

Researchers and developers building LlamaIndex applications that need citation-backed answers grounded in OpenAlex papers, including full-text search of Open Access works.

Source README

OpenAlex Reader

pip install llama-index-readers-openalex

This loader will search for papers in OpenAlex and load them in llama-index. The main advantage of using OpenAlex is that you can search the full-text for Open Access papers as well.

Usage

from llama_index.readers.openalex import OpenAlexReader

openalex_reader = OpenAlexReader(email="shauryr@gmail.com")
query = "biases in large language models"

### changing this to full_text=True will let you search full-text
documents = openalex_reader.load_data(query, full_text=False)

What can it do?

As shown in demo.ipynb we can get answers with citations.

query = "biases in large language models"
response = query_engine.query(
    "list the biases in large language models in a markdown table"
)
Output
Source Biases
Source 1 Data selection bias, social bias (gender, age, sexual orientation, ethnicity, religion, culture)
Source 2 Biases of what is right and wrong to do, reflecting ethical and moral norms of society
Source 3 Anti-Muslim bias
Source 6 Gender bias
Source 9 Anti-LGBTQ+ bias
Source 10 Potential bias in the output

FAQ

Common questions

Discussion

Questions & comments · 0

Sign In Sign in to leave a comment.