Hello Anahita tribe,
I have created the foundations for the com_hashtags and hashtagable behavior. So actors, mediums, and comments are hashtagable. Hashtags ought to be added to the body of the nodes to be extracted.
I have also developed a script that looks up all the instances of hashtag terms in the body of hashtagable nodes and creates actual hashtag records in the database. For now you can clone the Anahita fork from my account:
and run the migration:
php anahita db:migrate:up
to migrate up. This script is going to take a while to run. To give you an idea, for the Anahitapolis data it took 6 hours to run the script. Depending on the size of your data, you'd need to allocate a lot of memory to your php. To do that increase the memory_limit to 512MB or more. I tested the Anahitapolis data with 1024MB of memory limit so far and currently testing with 512MB. Let's see how much lower I can go.
After the script is finished a lot of hashtag nodes and graphs are added to the Anahita which can be used for all kinds of fun stuff and that is what I am going to work on next.
To see the hashtag data you can use the following queries:
Number of hashtag nodes:
SELECT COUNT(*) FROM jos_anahita_nodes WHERE TYPE LIKE '%hashtag%;
Number of hashtagable nodes:
SELECT COUNT(*) FROM jos_anahita_nodes WHERE hashtag_ids != '';
Number of edges connecting hashtags and hashtagables:
SELECT COUNT(*) FROM jos_anahita_edges WHERE TYPE LIKE '%hashtag%;
And to get a list of top 10 hashtags in your Anahita installation:
SELECT id, name, alias, hashtagable_count FROM jos_anahita_nodes WHERE TYPE LIKE '%hashtag%' ORDER BY hashtagable_count DESC limit 210;
that's it for now. Please create a local Anahita instance, feed the data from your live Anahita to it, and then run the migration. Please let me know of any errors, bugs, or issues.