Added performance testing results to readme.md (6 docs with openai with each graph database (neo4j, kuzu, falkordb)
Added docs/performance.md: has performance testing results for each graph database with 2,4,6 docs with openai and 2,4 docs with ollama
Added support for FalkorDB graph database https://www.falkordb.com/ and https://github.com/FalkorDB/falkordb The abstractions of LlamaIndex, LlamaIndex support for FalkorDB, and the configurability of flexible-graphrag made this a relatively straightforward process.
Added LlamaIndex DynamicLLMPathExtractor support (works on openai, not on ollama currently)
Added config of kg extractor type (simple, schema, or dynamic) to set which LlamaIndex extractor to use (SimpleLLMPathExtractor, SchemaLLMPathExtractor, or DynamicLLMPathExtractor)
Added config of MAX_TRIPLETS_PER_CHUNK and MAX_PATHS_PER_CHUNK
Added readme.md info on system environment setup of ollama for performance and parallelism (OLLAMA_CONTEXT_LENGTH, OLLAMA_NUM_PARALLEL, etc.)
Added new default schema with 35+ relationship combinations, more relations, and entity types: PERSON, ORGANIZATION, TECHNOLOGY, PROJECT, LOCATION
Fixed file upload dialog performance in all 3 front ends: React, Angular, and Vue (chosen files display quickly after dialog ok)
The Angular, React, and Vue frontend clients now have different stages organized into different tabs so they have room. They all can be switched between a dark and light theme using the slider at the top right corner. New functionality beyond the old UI includes a file upload dialog, drag/drop upload, a table with file processing progress bars, and a new Chat UI. Note the github readme.md page has collapse / expand sections to look at screenshots with dark and light themes for React, and only shows the light theme for Angular and Vue.
Sources Tab
Allows you to choose file to upload from the file system, or paths file or folder path in Alfresco or CMIS repositories. For filesystem files you can now use a file upload dialog and drag/drop files onto the drop area in the source tab view.
Here you can modify what files get processed by unselecting / selecting file checkboxes, Remove from processing list by using x a on file row, our use the remove selected button. The click on Start Processing to process selected files. There is an overall progress bar, and per file progress bars. Note currently all files are processed as one batch in the backend, so the file progress bars will be showing the same status. You can cancel processing by using the cancel button
Search Tab
Here you can do a Hybrid Search (Fulltext+Vector RAG+GraphRAG) or (Fulltext+Vector RAG) depending on configuration. This gives you a traditional results list. For now ignore the scores and extra results just check results order.
The Q&A Query, Here you ask a question using conversational style (This is an AI query using the configured LLM and the information submitted in the processing tab (and in full text, vector, and graph “memory”)
Chat Tab
This a traditional chat style UI allowing you the enter multiple conversational Q&A queries (AI queries like the one at a time in the Search Tab). You hit enter or click the arrow button to submit a query. You can also use Shift+Enter to get a extra new line for your question. The chat view area displays a history of questions and answers. The you can clear things with the Clear History button
Flexible RAG
I used Flexible RAG in the title to indicate that Flexible GraphRAG can be configured to just be a RAG system. This would still have the flexibility that LlamaIndex abstractions provide to be able to plug in different search engines/databases, vector databases, and LLMs. You still get Angular, React, and Vue frontends, have MCP server support, a FastAPI backend, and Docker support. You could just configure a search engine. You could just configure a Graph database for auto graph building of knowledge graphs using the configurable schema support.
For RAG configuration: Flexible GraphRAG can be setup to do RAG only without the GraphRAG (see env-sample.txt and setup your environment in .env, etc.):
Have SEARCH_DB and SEARCH_DB_CONFIG set for elasticsearch, opensearch, or bm25
Have VECTOR_DB and VECTOR_DB_CONFIG setup for neo4j, qdrant, elasticsearch, or opensearch
Have GRAPH_DB set to none and ENABLE_KNOWLEDGE_GRAPH=false.
Server Monitoring and Management UI
Basically you can use the docker setup and get a docker compose that run all the following at the same time (or a subset by commenting out a compose include) without having to these up individually: Alfresco docker compose (which has Share and ACA), Neo4j docker (which has a console URL), Kuzu API server (not used, used embedded), Kuzu explorer, Qdrant (which has a dashboard), Elasticsearch, Elasticsearch Kibana dashboard, OpenSearch which has a OpenSearch Dashboards URL.
So you can setup a browser window with tabs for all these dashboards, Alfresco Share / ACA, and Neo4J console. This is your monitoring and management UI.
You can uses the Neo4j, Elasticsearch Kibana, Qdrant dashboard, OpenSearch dashboards to delete full text indexes (Elasticsearch, OpenSearch), delete vector indexes (Qdrant, Neo4j, Elasticsearch, OpenSearch) and delete nodes and relationships (Neo4j and Kuzu consoles).
And the winner is using LLMs to create knowledge graphs over using NLP. Can LLMs do a better job? The Neo4j LLM Graph Builder in particular, has shown they can. What about the cost of using OpenAI along with the loss of privacy of data by submitting? The answer is free and local LLM models (Llama3 versions are available thru ollama) work too with Graph Builder. I tested with OpenAI GPT-4o, llama3, llama3.1, llama3.2. I noticed gemma2 is also available thru ollama. With these local LLMs, you will need a high end Nvidia card to work best.
The Online LLM Graph Builder can be used. You need to provide it with your Aura Neo4j connection info (you can create an account for a free Aura DB). It only has Diffbot, OpenAI, and Gemini LLM models available.
Graph Builder can upload from local files, AWS S3, web pages, Wikipedia, and Youtube. Google GCS can be a source if configured.
First choose the LLM model to use. Then upload one or more files. Then choose generate graph. You can view the graphs with the basic viewer (which allows hiding chunk nodes, community nodes, so you can see the entities and relationships). The Bloom viewer is also available, which is more complicated.
You can also chat with the data using GraphRAG and your chosen LLM. Answers have a icon below them that when clicked, provides info on graph doc sources, what entities, and what chunks were used to answer.
The online version doesn’t have the llama3 models. So you need to clone the github project and build locally. To add using Meta Llama3 models, you need to configure it. You use the example.env to create a .env file and then add an optional OpenAI key, LLM model configuration, and indicate you initial Neo4j database info. Neo4j connection info can also be provided in the UI. Then do docker compose up. I have a fork of the main branch in my LLM Graph Builder that has added: configuration for lllama3, llama3.1, llama3.2, and openai gpt-4 choices, some neo4j connection config examples, switched to 8090 to not conflict with Alfresco 8080, has an additional debug log to so you can check on model config. and has a sample files folder with space-station.txt.
Speaking of Alfresco, I could add to my Alfresco GenAI Semantic project to call the separable backend of Graph Builder to generate a knowledge graph of new or updated Alfresco documents that have a new custom aspect. The backend may only have support for sources coming for the app’s kinds of sources currently. Also note in terms of UI integration, Alfresco’s ADF components and the ACA client use Angular. Neo4j Graph Builder’s front end uses React (and so does some of their other software projects).
space-station.txt with OpenAI GPT-4o:
space-station.txt with Meta Llama3:
space-station.txt with Meta Llama3.1:
space-station.txt with smaller Meta Llama3.2:
OpenAI GPT-4o with Albert Einstein Wikipedia page (340 nodes, 230 relationships):
Meta Llama3 with Albert Einstein Wikipedia page (150 nodes, 150 relationships), not shown: Llama3.1 (had 161 nodes, 85 relationships), not shown Llama3.2 (125 nodes, 76 relationships)
The Alfresco GenAI Semantic github project now adds regular Alfresco tags when performing auto tagging when enhancing with links to Wikidata and DBpedia. Semantic entity linking info is kept in 3 parallel multi-value properties (labels, links, super type lists) in the WikiData and DBpedia custom aspects. The labels values are used for the tag labels.
I switched to a local, private Wikidata recognizer. The spaCy-entity-linker python library is used for getting Wikidata entity links without having to call a public serivce api. It was created before spaCy had its own entity linking system. It still has the advantage of not needing to do training. Had previously used the spaCyOpenTapioca library, which calls an OpenTapioca public web service api URL. Note the URLs in the links properties do go to the public website wikidata.org if used in your application.
I also switched to a local, private DBpedia Spotlight entity recognizer in a docker composed in. The local URL to this docker is given the to the spacy DBpedia Spotlight for SpaCy library. This library was using a public Spotlight web service api URL by default previously. Note the URLs in the links properties do go to to the public website dbpeda.org if used in your application.
For documents with the Wikidata or DBpedia aspects added to them, tags will show up in the Alfresco clients (ACA, ADW, Share) after PDF rendition creation and alfresco-genai-semantic AI Listener gets responses from REST apis in the genai-stack. Shown below are tags in the ACA community content app:
Multi-value Wikidata aspect properties of a document in the ACA client are shown below in the view details expanded out. The labels property repeats what the labels of the tags have. The links properties have URLs to wikidata.org. The super types properties have the zero “” or one or multiple comma separated super types in wikidata for each entity. These supertypes are wikidata ids (are links once you add “http://www.wikidata.org/wiki/” in front of the ids).
The same style DBpedia aspect multivalue properties are shown below in the ACA client. Note that the super types can be from Wikidata, DBpedia, Schema (schema.org), foaf, or DUL (ontologydesignpatterns.org DUL.owl), etc.
The Alfresco GenAI project provides support for generative AI with local or cloud LLMs for Alfresco. This includes summarization, categorization, image description, chat prompting about doc content.
The Alfresco GenAI Semantic project adds named entity recognition (NER) / entity linking of documents in Alfresco to Wikidata and DBpedia. Currently 2 custom aspects have multi-value properties for the links, alfresco tags aren’t used yet.
The spaCy NLP python library along with spaCy projects are used. The spaCyOpenTapioca project is used for getting Wikidata entity links. The DBpedia Spotlight for SpaCy project is used for getting DBpedia entity links. Note these both use external servers, which can be setup locally. NER can also be done with just spaCy. The spaCy-LLM python package that integrates Large Language Models (LLMs) into spaCy pipelines is available. The Alfresco GenAI Semantic project currently doesn’t use spacy-llm yet.
Below shows what test\space-station.txt after upload and entity linking with the Entity Link Wikidata aspect looks like in the Alfresco ACA content app in the view details when expanded out:
Below shows what test\space-station.txt after entity linking with the Entity Link DBpedia aspect looks like in the Alfresco ACA content app in the view details when expanded out:
