Skip to content
This repository was archived by the owner on Mar 19, 2019. It is now read-only.

The Database Model

Jump to bottom
jbytecoder edited this page May 21, 2015 · 5 revisions

The project will manage data described in this data model: DB Data Model

Some explanation about diagram conventiions. It represents Entities as rectangles, and their properties as ovals. Name attributes are obvious, Date property in every entity represents date and time when the data was added. COunt is also obvious. Share property needs definiotion. Share is a precentage. It represents the share of the Category in total count of detected keywords. Entries form particular date should sum up to 100%

Red entities are designed to be writable by sprak and readonly by forntned. Green entities are qutie opposite. The yellow entitites are designed to be writable in spark, but deletable from fronted and read by frontend

The idea behind this model is as folows. Sprak queries latest tweets. Based on this it populates the UserActivity, and TagActivity with any tags and users it encounters if new entities are encuntered it fills the User and HashTag tables. This is most basic sprak dutey performed periodically. Now the control goes to frontend and user.

By means of Forntend user chooses users and tags that are to be traced. The fronted populates MonitoredUser and MonitoredHashTag. User allso defines the categories and keywords that will be sought for.

Sprak should deteck that theese data are in place. When that happens it should start processing latest tweets of monitored users periodically, looking or counting occurences of keywords. Spark should then count the precentage of keyword uses in all present categories.

This ERD Model provides bit deeper description of data model All attributes are not null, underlined attributes are uniqe ERD

Clone this wiki locally