Diego Esteves

Pickles have been converted to Python3 via conversion script, this _might_ be a problem in terms of correctness (end-to-end results). All pickles should be retrained from scratch using now Python3.

DB local file is not scalable, we need to perform queries to Elasticsearch instead. 1. set up a local instance of ES 2. migrate data from DB to ES. 2.1...

I have found an issue that might be affecting the performance of the NER model. I am currently using a sort of snippet of the webpage (from the Microsoft search...

We need to deploy a UI to **visually** validate the results and check what is being returned from the search engine, otherwise, we are either blind (just look to overall...

CRF layer performance

2

Compare standard feature function x dynamic feature buider. There is still a `0.1` dif in `F1`. Check why! `Notebook src/training/notebooks/horus_v1/03-horus-training-ner-crf.ipynb`

https://github.com/leondz/entity_recognition

https://duckduckgo.com/api

Index Web Data Commons (CW12)

1

http://webdatacommons.org/structureddata/2017-12/files/html-rdfa.list ```horus/core/search_engines.py``` and ```horus/core/service.py``` files.

```horus/core/search_engines.py``` and ```horus/core/service.py``` files.