Comments (2)
A quick analysis shows no bug. If you set a 0 < bias < 1 it will increase the scores over removing the fields and therefore having no match at all for that field because but because you are adding a new match, which is additive you see some small increase. In other words the fact of adding the field will increase over no field but a bias <1 will make the match arbitrarily small, approaching 0, which would be equal to omitting the field. Another way to say this is that the deboost is in comparison to any other form of bias.
So given the presentence of the field < 1 is deboosting.
Anyway this is as designed. i suppose what you want is a new feature, which is to move the field to the "should_not" clause of the ES query with it's own bias?
from universal-recommender.
Let me try to explain the deboost better.
Imagine 2 fields: housewares, and electronics. If you want to boost electronics and deboost housewares you would apply bias = 1.5 to electronics and 0.5 to electronics. then houseware will be ranked higher that electronics.
Another thing you can do is apply a boost of 0 to housewares and none will be shown, they will all be excluded.
I don't think these work for your needs so understand the issue but anyway, it's the way things work today.
from universal-recommender.
Related Issues (20)
- java.lang.NegativeArraySizeException error for secondary indicators HOT 3
- Error during training when using remote ElasticSearch HOT 20
- UR is using elastic even when all the data sources are MYSQL HOT 1
- would you help me solve the problem: "had a not serializable result: org.apache.mahout.math.DenseVector"
- Training Error - "Extracting datasource params. No 'name' is found. " HOT 1
- Location based recommendations HOT 2
- Cannot detect GPU
- Cannot get the recommender queries? HOT 3
- How to "Delete" items ? HOT 1
- Error on training HOT 1
- Include filter rule returning empty list HOT 1
- recommendations is empty. Training job status is failed. But in harness log, everything is ok
- TTL resets to 365 on every restart
- Can not train data
- Training data occasionally has errors HOT 1
- README documentation outdated
- The $delete event does not work
- Hello, I got this error when I try to train the model HOT 3
- business rule with contains instead of exact match
- Do anyone have Python version of this project? I don't know Scala
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from universal-recommender.