Anonymize Documents, Store Information, and Access via Web Services

**Project for Document Anonymization, Information Storage, and Access via Web Services** 1. **Types of Input Documents:** - The system must accept and process the following types of documents: - Images (specify accepted formats, e.g., JPG, PNG, etc.). - PDF. - Word Documents (.docx). - Plain Text Files (.txt). - String as parameter 2. **Detection and Anonymization of Sensitive Data:** - Implement the detection of sensitive data (names, addresses, identification numbers, etc.) and their anonymization using SpaCy or another technology that offers better performance and accuracy. The solution must be capable of handling various types of sensitive data and should be easily configurable for new data categories if needed. 3. **Storage in Elasticsearch:** - All processed document information must be stored in Elasticsearch with the following fields: - **Identification Number**: Provided through the API with each request. - **Source Document Name**: Including the extension. - **Source Document Path**: To access the original file. - **Anonymization Date**: To track when the document was processed. - **Anonymized Text**: The resulting text after anonymization. - **Subject Type (optional)**: A field to categorize the type of subject matter of the document. - **Summary (optional)**: A summary of the document’s content. - The combination of the identification number and the file name will make each record in Elasticsearch unique. 4. **Processing of Multiple Documents:** - The system must support the processing of up to 10 documents in a single request. A single API request may include the names and paths of up to 10 files, which should be processed and stored as separate records in Elasticsearch but linked by the same identification number. 5. **API in Flask:** - Create a RESTful API in Flask that offers the following functionalities: - **Start Anonymization**: It should allow the anonymization of up to 10 specified files in a single request. The subject type and summary fields will be optional in this function, allowing them to be included directly if this data is already available. - **Data Query**: It should include endpoints to query anonymized documents stored in Elasticsearch. - The API must be accessible from external applications, especially from a PHP application. - Include an integration example with PHP to demonstrate the API's capability, ensuring effective interaction with it from other environments. **Additional Comments:** - **Request Management**: Ensure the API correctly handles requests with multiple files, processing each independently while maintaining the link to the provided identification number. - **Unique Identification**: Implement mechanisms to ensure that the combination of identification number and file name is stored uniquely in Elasticsearch, preventing duplicates and integration errors. **Data to be Anonymized:** 1. Names of people and organizations, legal document identifiers, addresses. 2. There is no template for source documents; there are many types of documents. Anonimized data must be substituted by contacto1, contacto2, y so on for each person or organization, id1,id2 and so on for id of documents, direccion1, direccion2,’ and so on for addresses **Consider 4 hours to help us configure our server and explain the necessary configuration.** If you plan to use any tool that requires a license, you must specify which one and its cost in the proposal.