Actions
Feature #1036
closedSearch for similar names
Start date:
2019-06-07
Estimated time:
16.00 h
Description
To prevent duplicates or spelling mistakes a search for similar names will be implemented.
- Adding Python library fuzzywuzzy which uses Levenshtein Distance to calculate the differences between sequences
- Option to select ratio
- Option to select class
Ideas for next version
- Search for a manual added string
- Add a check when inserting an entity and warn if a similar name already exists
Updated by Alexander Watzinger over 5 years ago
Or maybe use PostgreSQL with install postgresql-contrib:
CREATE EXTENSION pg_trgm;
CREATE INDEX trgm_idx ON model.entity USING GIST (name gist_trgm_ops);
select (similarity(n1.name, n2.name)) as sim, n1.name, n2.name
from model.entity n1, model.entity n2
where
n1.id != n2.id
and n1.system_type = 'place'
and n2.system_type = 'place'
and similarity(n1.name, n2.name) > .7
order by sim desc;
Updated by Alexander Watzinger over 5 years ago
- Description updated (diff)
- Status changed from In Progress to Closed
Actions