Step 1. Training the Model
$ ./train.sh ur_default
Step 2. Building the Index
$ ./bin/sqlcco_index_build --sqlite sqlcco.db --fts sqlcco.json.gz
Step 3. Deploying the Model
$ ./deploy.sh
Step 4. Querying the Model
$ ./bin/sqlcco_recommend --sqlite sqlcco-deploy.db --fts sqlcco.json.gz --entity u1
Step 5. Serving the Model
$ ./bin/sqlcco_serve --sqlite sqlcco-deploy.db --fts sqlcco.json.gz --address 0.0.0.0 --port 8080
Step 6. Querying the Model (via HTTP)
$ curl -s 'http://127.0.0.1:8080/recommendation/u1?limit=20&normalized=true' | jq
Update Entity History
New user browses electronics
$ curl -s 'http://127.0.0.1:8080/log/u5?indicator=category-browse&target=electronics'
New user sees phones
$ curl -s 'http://127.0.0.1:8080/log/u5?indicator=category-browse&target=phones'
New user views iphone
$ curl -s 'http://127.0.0.1:8080/log/u5?indicator=view&target=iphone'
Additional
Searching the Index (via HTTP)
$ ./bin/sqlcco_search --sqlite sqlcco-deploy.db --fts sqlcco.json.gz --query '{"gender":"female","country":"US","language":"en","age":"21"}'
Searching the Index (via HTTP)
curl -s -H 'Content-Type: application/json' -d '{"gender":"female","country":"US","language":"en","age":"21"}' http://127.0.0.1:8080/search | jq
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent