Optimize Document Retrieval with Jina Reranker Cognitive Actions

22 Apr 2025
Optimize Document Retrieval with Jina Reranker Cognitive Actions

The Jina Reranker from the spec yxzwayne/jina-reranker-v1-turbo-en enables developers to efficiently retrieve and rank the most relevant documents from a specified list based on a given query. This action is particularly useful in applications that require high accuracy in search operations while being mindful of performance. By leveraging this pre-built action, developers can save time on complex search algorithms and enhance user experience through improved document relevance.

Prerequisites

To utilize the Jina Reranker Cognitive Action, you will need an API key for the Cognitive Actions platform. This key is essential for authentication, typically passed in the headers of your requests. Ensure you have your environment set up to make HTTP requests, and you are ready to integrate the Jina Turbo Reranker into your application.

Cognitive Actions Overview

Perform Jina Turbo Reranking

The Perform Jina Turbo Reranking action executes a highly efficient Jina Turbo Reranker to retrieve and rank documents based on the user's query. It allows you to define how many of the top results you want to receive, optimizing both accuracy and performance.

Input

The input for this action is structured as follows:

  • documentList (required): A stringified list of documents within which the search will take place. Each document must be represented as a string.
  • query (required): The search query string used to identify the most relevant documents.
  • topK (optional): Specifies the number of documents to return, defaulting to 5. This parameter determines how many top documents the query should retrieve.

Example Input:

{
  "topK": 3,
  "query": "should we always bet on the attention mechanism?",
  "documentList": "[\"state space models remain robust for time series analysis without requiring such complex structures.\", \"Variational inference offers an efficient way to approach complex probabilistic models, which can sometimes be more advantageous.\", \"Always relying on attention mechanisms may not be ideal for every AI application, as it can lead to increased computational costs and overfitting in certain scenarios.\"]"
}

Output

The action returns a list of the most relevant documents along with their corresponding relevance scores. Each output entry contains:

  • text: The document text.
  • score: A floating-point number representing the relevance score of the document.
  • corpus_id: An identifier for the document in the corpus.

Example Output:

[
  {
    "text": "state space models remain robust for time series analysis without requiring such complex structures.",
    "score": 0.5842115879058838,
    "corpus_id": 0
  },
  {
    "text": "Always relying on attention mechanisms may not be ideal for every AI application, as it can lead to increased computational costs and overfitting in certain scenarios.",
    "score": 0.4497090280056,
    "corpus_id": 2
  },
  {
    "text": "Variational inference offers an efficient way to approach complex probabilistic models, which can sometimes be more advantageous.",
    "score": 0.28855061531066895,
    "corpus_id": 1
  }
]

Conceptual Usage Example (Python)

Here’s how you might call the Jina Turbo Reranker action using Python:

import requests
import json

# Replace with your Cognitive Actions API key and endpoint
COGNITIVE_ACTIONS_API_KEY = "YOUR_COGNITIVE_ACTIONS_API_KEY"
COGNITIVE_ACTIONS_EXECUTE_URL = "https://api.cognitiveactions.com/actions/execute" # Hypothetical endpoint

action_id = "95285003-fd33-4406-a8c0-34098b373bd4" # Action ID for Perform Jina Turbo Reranking

# Construct the input payload based on the action's requirements
payload = {
    "topK": 3,
    "query": "should we always bet on the attention mechanism?",
    "documentList": "[\"state space models remain robust for time series analysis without requiring such complex structures.\", \"Variational inference offers an efficient way to approach complex probabilistic models, which can sometimes be more advantageous.\", \"Always relying on attention mechanisms may not be ideal for every AI application, as it can lead to increased computational costs and overfitting in certain scenarios.\"]"
}

headers = {
    "Authorization": f"Bearer {COGNITIVE_ACTIONS_API_KEY}",
    "Content-Type": "application/json"
}

try:
    response = requests.post(
        COGNITIVE_ACTIONS_EXECUTE_URL,
        headers=headers,
        json={"action_id": action_id, "inputs": payload} # Hypothetical structure
    )
    response.raise_for_status() # Raise an exception for bad status codes (4xx or 5xx)

    result = response.json()
    print("Action executed successfully:")
    print(json.dumps(result, indent=2))

except requests.exceptions.RequestException as e:
    print(f"Error executing action {action_id}: {e}")
    if e.response is not None:
        print(f"Response status: {e.response.status_code}")
        try:
            print(f"Response body: {e.response.json()}")
        except json.JSONDecodeError:
            print(f"Response body: {e.response.text}")

In this code snippet, you replace the placeholder for the API key and the execution URL. The action ID for the Jina Turbo Reranker is specified, and the input payload is structured according to the requirements of the action.

Conclusion

The Jina Reranker Cognitive Actions provide an effective tool for developers looking to enhance document retrieval and ranking capabilities in their applications. By leveraging this action, you can streamline the search process, ensuring users receive the most relevant results with high efficiency. Next, consider integrating this functionality into your search applications to improve user satisfaction and engagement.

← Back to All Articles