DIG search UI not getting populated with events sample project

[sobalgi]

Hello Mr. Singh,

I have followed the instructions for hands-on example project 'events' as mentioned in the advanced user guide. I'm not able to obtain the expected DIG Search UI after running extraction pipeline using 'Recreate Knowledge Graph'. I get a blank UI

I have the KG entries fully populated

I get the following logs in the terminal when the engine is running Is the ES instance running fine or is the logstash unable to extract logs because of this error ? Also " high disk watermark [90%] exceeded on [zlPx-2GRS6WsA0cR_D131w][zlPx-2G][/usr/share/elasticsearch/data/nodes/0] free: 7.2gb[3.1%], shards will be relocated away from this node" is being showed on the terminal. Is there a chance that this is causing the problem ?

I have defined the supplementary ETK config file as indicated in the instructions. Firstly while creating 'etk_config_supplement.json' file, there was no folder named 'additional_etk_config'. There were only 'additional_ems', 'generated_additional_ems' and 'generated_em' folders. So I created the 'additional_etk_config' folder myself and later 'etk_config_supplement.json' file in it.

Also I can see a change in the DIG UI after the 13th September commits. Are the user instructions still valid for the newer UI? As I understand the user guides are not updated regularly.

I request your assistance to get sample project up and running so that I can proceed further. Many thanks in advance. Best regards, Sourabh.

[saggu]

The UI does not show you any documents by default but it does tell you there are 200 docs in the knowledge graph.

If you click on the spanner you can search in the individual fields.

You are right about the user documentation not being updated. We shall update with for the new UI and the way myDIG works internally.

Thanks ,

[sobalgi]

Many thanks for the quick response. I really appreciate it. Regarding the ES instance and logstash error as previously mentioned, is that the expected behavior ?

If I click on the spanner and next the search button, I don't see any textbox for searching.

Also the 'additional_etk_config' folder not already being present is not a problem I understand. Is it ?

The Hands-on instructions so far are excellent. Even though there is a slight change in gui it can be figured out easily. Additional instructions on creating the custom extractors for fields would be very helpful as this would constitute the core of extracting entities and populating the KG.

A short instruction on using Kibana would also be very helpful.

Many thanks, sobalgi.

[saggu]

There is no need to have the folder additional_etk_config as we do not extract using etk configs anymore. Take a look at etk examples - https://github.com/usc-isi-i2/etk

Now we write etk modules instead of configs.

In regards to not getting any fields to search in, please post a screenshot when you click on the spanner.

We are working on updating the documentation.

[sobalgi]

ok. Thanks for the etk link. I'll go through that. So we don't need the additional config json file now.

The screenshot of the spanner to search fields. As you can see there are 200 events documents. But the facets in the left are not populated and neither is the search window.

It would be very helpful to have the updated documentation and hands on guide.

Many thanks.

[saggu]

Do you have any fields in myDIG ? If so, edit those fields which you want to search on and check show in search

[sobalgi]

Awesome!!! Now I can see the fields.

As a next step, I tried searching for a location in the search fields as it's present in the document given below.

I tried searching for Faridpur,Bangladesh in the location as I have already confirmed it to be present in the .jl file with 'doc_id : 0aa32939-bae6-11e7-af3a-cc2f7122e7f0'.

But I get no search results even though the keywords are present in KG.

Can you please confirm this?