Hello! Dr. Yu Wang ~ I'm currently working on running the KG-LLM-MDQA project found in pipeline_demo.zip and have come across some technical challenges. I believe this part of the project to be the MVP, encompassing graph construction to graph retrieval, and I'm keen on getting it to work properly.
Issue: ModuleNotFoundError in backend.py
While attempting to run backend.py, I'm encountering a ModuleNotFoundError. The error details are as follows:
I have checked the directory structure but can't seem to locate the related module or file. This suggests that some files might be missing, or perhaps I missed some crucial setup steps. Could you provide guidance or suggestions for resolving this?
Database Selection Considerations
Additionally, I'm contemplating the choice of database for storing extensive text segments used in the project. Considering these text segments as nodes in the database, I'm evaluating the appropriateness of Neo4j for this purpose. My main considerations are:
Search Efficiency: How well does Neo4j perform in handling large volumes of text data, especially considering the need for rapid node (text segment) search in recommendation flows?
Edge Management: Can Neo4j adeptly manage dynamically changing relational data, given that edge updates and queries might be more frequent than nodes?
Distributed Data Consistency: In distributed systems, managing data consistency is crucial. How effectively can Neo4j handle this in a distributed environment?
Feasibility for Text Storage: While Neo4j excels with relational data, is it the best option for storing substantial text segments?
I'm also weighing the option of using a combination of Elasticsearch and MongoDB. Could you provide insights on which approach might be more suitable for this kind of data handling?
Thank you for your time and assistance. I'm very enthusiastic about your project and look forward to making the most out of it.
Hello! Dr. Yu Wang ~ I'm currently working on running the KG-LLM-MDQA project found in pipeline_demo.zip and have come across some technical challenges. I believe this part of the project to be the MVP, encompassing graph construction to graph retrieval, and I'm keen on getting it to work properly.
Issue: ModuleNotFoundError in backend.py While attempting to run backend.py, I'm encountering a ModuleNotFoundError. The error details are as follows:
I have checked the directory structure but can't seem to locate the related module or file. This suggests that some files might be missing, or perhaps I missed some crucial setup steps. Could you provide guidance or suggestions for resolving this?
Database Selection Considerations Additionally, I'm contemplating the choice of database for storing extensive text segments used in the project. Considering these text segments as nodes in the database, I'm evaluating the appropriateness of Neo4j for this purpose. My main considerations are:
Search Efficiency: How well does Neo4j perform in handling large volumes of text data, especially considering the need for rapid node (text segment) search in recommendation flows?
Edge Management: Can Neo4j adeptly manage dynamically changing relational data, given that edge updates and queries might be more frequent than nodes?
Distributed Data Consistency: In distributed systems, managing data consistency is crucial. How effectively can Neo4j handle this in a distributed environment?
Feasibility for Text Storage: While Neo4j excels with relational data, is it the best option for storing substantial text segments?
I'm also weighing the option of using a combination of Elasticsearch and MongoDB. Could you provide insights on which approach might be more suitable for this kind of data handling?
Thank you for your time and assistance. I'm very enthusiastic about your project and look forward to making the most out of it.