Open RanAR90 opened 6 years ago
Sorry for the late rely. I haven't used this account for quite a while. I hope you had figure out the issue. In short, the database was used for storing relation triplets. I think you case was that either the algorithm did not find any relation teiplets (you can print them out) or the triplets were not saved to database. Also, I have been very busy and will be busy till July or August. I will only have time to write a documentation or some simple examples one or two months later.
Sorry, I still do not have time to write the example. I have to push it towards the New Year when I have a break. You could try to make the hierarchical topic modeling work first. One way to debug it was to print out all the relation triplets you extracted from texts. The code saves the triplets to a database, and the records are later used to add properties and to the RDF ontology.
I don't know if there is a way to push a database example without showing the database name or username password. This should be straightforward anyway. hrLDA/src/com/hrLDA/mysql/MySqlDB.java defines the schema/structure of the tables. Will it help if I just post some screenshots of the command line outputs?
Hello..
I would like to thank you for the very great work. It is very useful, many thanks for making it available to other researchers. I am trying to test it against my corpus of text, but getting this error message "no data for the seeds" when trying to initialize the HierarchicalRLDA. am I supposed to extract the relations using ollie my self and load them into the database we are setting in the parameters file? For now, I am using an empty database as I thought these relations will be extracted and inserted into the database in the run-time. I can see that the instances passed contain '0's only in the sentences and weights and I feel I am missing something here (if an example of the database structure is available, it would be very much appreciated if you can guide me where to find it). Can you please advice?
Thanks again for your appreciated work. Very kind regards