Investigating Independent Data Processing for Custom Data Querying

Anonymous

Abstract: Large Language Models (LLMs) have become increasingly valuable for querying personal data, including structured and unstructured information. However, these models face limitations due to the context window size, which restricts the inclusion of multiple lengthy documents. To overcome this constraint, we provide the investigation of a technique for custom data querying, that enables the independent processing of individual data. This approach enhances control over data processing, simplifies data manipulation, and increases the number of data items that can be handled. Our empirical results highlight the validity of this method for various data types, showing promising outcomes for applications involving personal data retrieval. This paper introduces a valuable approach for expanding the capabilities of LLMs in personalized data querying.

Paper Type: short

Research Area: Efficient/Low-Resource Methods for NLP

Contribution Types: Model analysis & interpretability, NLP engineering experiment

Languages Studied: english

Consent To Share Submission Details: On behalf of all authors, we agree to the terms above to share our submission details.