# Stateful conversation graph **URL:** \ **Category:** Projects & Collaboration\ **Tags:** neo4j-desktop\ **Created:** [October 15, 2020, 5:46am UTC](https://community.neo4j.com/t/stateful-conversation-graph/27245 "2020-10-15T05:46:31Z")\ **Posts on this page:** 9\ **Page:** 1
**Author:** ![varshamr96](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/varshamr96/32/14759_2.png) [@varshamr96](https://community.neo4j.com/u/varshamr96)\ **Post date:** [October 15, 2020, 5:46am UTC](https://community.neo4j.com/t/stateful-conversation-graph/27245/1 "2020-10-15T05:46:31Z")
1. Create a statefulness between a conversation 2. Link the conversation with a context and build a graph database 3. The example of the conversation is: 4. What is the covid status? 5. How many are hospitalized in Bangalore? 6. What is the recovery rate? 7. When the vaccine will be available? 8. This should create a graph with COVID as node and rest of the nouns as relationship ---
**Author:** ![tony.chiboucas](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/tony.chiboucas/32/5980_2.png) [@tony.chiboucas](https://community.neo4j.com/u/tony.chiboucas)\ **Post date:** [October 16, 2020, 10:08pm UTC](https://community.neo4j.com/t/stateful-conversation-graph/27245/2 "2020-10-16T22:08:53Z")
You're talking Conversational Design. Very tricky subject, for which a graph database is uniquely suited. Something like [Hume](https://graphaware.com/products/hume/) language processing would take you a long way in the right direction. After that, it's about identifying the speaker's intent, and connecting that to a response which ideally prompts for a more specific inquiry. ---
**Author:** ![varshamr96](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/varshamr96/32/14759_2.png) [@varshamr96](https://community.neo4j.com/u/varshamr96)\ **Post date:** [October 31, 2020, 2:55pm UTC](https://community.neo4j.com/t/stateful-conversation-graph/27245/4 "2020-10-31T14:55:23Z")
How go eshtablish the relationship to one conversation to another conversation And connect 2 differe graphdatabases with relationship ![IMG_20201029_104802](https://us1.discourse-cdn.com/flex021/uploads/neo4jcommunity/original/2X/4/423af548d245ee1f356ef109de9b03c2e796fee3.jpeg) ---
**Author:** ![tony.chiboucas](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/tony.chiboucas/32/5980_2.png) [@tony.chiboucas](https://community.neo4j.com/u/tony.chiboucas)\ **Post date:** [November 2, 2020, 6:09pm UTC](https://community.neo4j.com/t/stateful-conversation-graph/27245/5 "2020-11-02T18:09:07Z")
Natural language processing is not as simple as making relationships between conversations. Instead, it's about making a graph of each word, and it's meaning, along with the context of the conversation. From that, your graph would have a great deal of "knowledge" about any given statement or question. > **[Natural Language Processing (NLP) - Developer Guides](https://neo4j.com/developer/graph-data-science/nlp/)** > > This guide explains the available tools for doing Natural Language Processing (NLP) with Neo4j. Once you have that graph, you could _then_ build into it a conversation interface that used the graph, and the speaker's query, to extract the speaker's intent, and match that to the desired response or an inquiry for more specifics. However, crafting this into a naturally feeling interaction between the speaker and a computer will require careful [conversational design](https://designguidelines.withgoogle.com/conversation/conversation-design/what-is-conversation-design.html#what-is-conversation-design-what-isnt-conversation-design) ---
**Author:** ![varshamr96](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/varshamr96/32/14759_2.png) [@varshamr96](https://community.neo4j.com/u/varshamr96)\ **Post date:** [November 9, 2020, 9:59am UTC](https://community.neo4j.com/t/stateful-conversation-graph/27245/6 "2020-11-09T09:59:30Z")
Sir one doute Any other way to combine 2 conversational graph database automatically And also how to combine 2 graph database into one graph database ---
**Author:** ![tony.chiboucas](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/tony.chiboucas/32/5980_2.png) [@tony.chiboucas](https://community.neo4j.com/u/tony.chiboucas)\ **Post date:** [November 9, 2020, 6:58pm UTC](https://community.neo4j.com/t/stateful-conversation-graph/27245/7 "2020-11-09T18:58:31Z")
> [@varshamr96](#): > > Any other way to combine 2 conversational graph database automatically Building automation takes more effort than doing something manually. So, yes, there are ways to do that automatically, but you'd still have to write the Cypher to do so. > [@varshamr96](#): > > And also how to combine 2 graph database into one graph database If you're using Neo4j Desktop, you can add both dbs to the Neo4j instance, then [query across both databases](https://neo4j.com/developer/multi-tenancy-worked-example/#querying-across-databases) to move the data around. > **[Managing Multiple Databases in Neo4j - Developer Guides](https://neo4j.com/developer/manage-multiple-databases/)** > > In this guide, we will walk through how to manage multiple database within the Neo4j DBMS. ---
**Author:** ![varshamr96](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/varshamr96/32/14759_2.png) [@varshamr96](https://community.neo4j.com/u/varshamr96)\ **Post date:** [November 10, 2020, 2:00pm UTC](https://community.neo4j.com/t/stateful-conversation-graph/27245/8 "2020-11-10T14:00:18Z")
hlo sir gud evng . how to combine 2 sentences automatically based on the context in single graph database programmatically ---
**Author:** ![varshamr96](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/varshamr96/32/14759_2.png) [@varshamr96](https://community.neo4j.com/u/varshamr96)\ **Post date:** [November 10, 2020, 2:10pm UTC](https://community.neo4j.com/t/stateful-conversation-graph/27245/9 "2020-11-10T14:10:52Z")
hello sir any tool is their to build a conversation graph . ---
**Author:** ![tony.chiboucas](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/tony.chiboucas/32/5980_2.png) [@tony.chiboucas](https://community.neo4j.com/u/tony.chiboucas)\ **Post date:** [November 10, 2020, 5:35pm UTC](https://community.neo4j.com/t/stateful-conversation-graph/27245/10 "2020-11-10T17:35:36Z")
hlo sr, graph data need u 2 git gud