BlueBrain / BlueBrain/nexus

Studio: Dashboard Creation not end-user friendly

Open
#3,303 1 comment 0 reactions 0 assignees View on GitHub
feature fusion fusion-studio
Dominant language
Scala
Stars
301
Forks
76
PR merge metrics
No merged PRs in 30d

Description

**Is your feature request related to a problem? Please describe.**

The current way a dashboard is created is not user friendly. A very large knowledge of details in how you search the graph is required to create a meaningful dashboard (sparql, elasticsearch, ...).

I tried to follow the example and description given in the tutorial [HERE](https://bluebrainnexus.io/docs/getting-started/try-nexus.html#step-4-organize-and-visualize-cells-in-nexus-fusion) but I failed to understand how the search is constructed.

I finally was able to create a search, but the only entities I found were the dashboards I created in that studio. That left me with a lot of confusion, as thy I was not able to see any other entities. Maybe I have to search other entities first and to "import" them to my project or studio, after which I am able to search them in the studio? That is completely unclear and very confusing.

In addition, I do not see any description on the different views that I can select here.

Also in addition, when you click on "preview" it looks like you have to define the column type. I had the impression the data in nexus is kind of "organized", so it is clear: this data is a name, hence text. This data is a creation date, hence the column type is "date". I do not understand at all why the user has to specify the data type here.

An example search I have tried:

```
SELECT DISTINCT ?self ?s WHERE {
?s nxv:self ?self;
a nsg:Studio.
} LIMIT 20
```

**Describe the solution you'd like**

1. Make it simple. Pre-select a normal standard view, or better remove that dropdown box entirely from the "Create Dashboard".
2. Make it simple. Give many different examples of specific SPARQL queries a user can use in her/his dashboard. Easy copy-and-paste example for some basic usecases a user might have, with clear and detailed explanation of what the search query is doing.
3. Make it simple. Make sure the user will see some data!
4. Intrinsically define the column type.

![Screenshot 2022-05-27 at 11.03.22.png](https://images.zenhubusercontent.com/62714e29b18ff51479a15c78/97281e99-52c9-4c92-a9b3-35b02ff723f5)

![Screenshot 2022-05-27 at 11.04.24.png](https://images.zenhubusercontent.com/62714e29b18ff51479a15c78/3456251d-8d84-4132-b384-9f4e370f2c2e)

Contributor guide

Open the contributing guide

Research direction

Start with the Studio Create Dashboard flow and the linked tutorial, then review how the selected view, SPARQL search, preview, and column types are presented. Done would require a simpler creation flow, copy-and-paste SPARQL examples with explanations, visible data for a basic search, and automatic column-type handling.

Written by the indexing model from the issue text.

Assessment

Tech stack
elasticsearch, scala
Domain
data-visualization, frontend, search
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.