School Violence and Elastic

We are a group of volunteers ( who want to make a cool idea real and valuable. It’s about protecting all our kids. Problem—we do not know Elastic as well as we wish. So here’s the challenge: we want to eventually create a national database to reduce school violence. We want to start with a proof of concept app—using a 100K datasets. Here’s where you can help. CAN THIS BE DONE WITH ELASTIC AND ITS OTHER TOOLS?

We want school personnel, parents, researchers, mental health professionals, and policy analysts to be able to use a web app that functions like this:

1st Use

  1. user types in observations made about a student in text fields:
    dressing differently
    break up with girlfriend
    single-parent home
    lowering grades
    new car
    more withdrawn

  2. hits submit button

  3. yields % of those terms involved in the following Department of Justice , CDC, and NIH recognized forms of school violence—like this

71% of drugs brought to campus
23% of vandalism
33% of sexual harassment
44% of assaults
85% of bullying
19% of stealing
10% of bringing weapons to campus
44% of verbal abuse

2nd Use
Automatic Extended Search
App also volunteers other behavior terms that provide additional insight or high percentages:

Added Terms: When “ten miles from school,” “leather coat,” and “family income 25K” are added, the following changes occur:

Results: Terms Appear Related in Following Categories of Violence
87% of drugs brought to campus versus 71%
28% versus 23% of vandalism
33% versus 33% of sexual harassment
56% versus 44% of assaults
89% versus 85% of bullying
21% versus 19% of stealing
67% versus 10% of bringing weapons to campus
67% versus 44% of verbal abuse
number of searched records: 8, 243, 545

3rd Use –User Gets Percentages and then Queries by
Zip Code
Other demographics to be entered later (ie. type of family, income etc.)

4th Use Searches by number of queries made—for use by school resource officers and police by

Zip Code
Other demographics to be entered later (ie. type of family, income etc.)

5th Use – Searches by terms by type of user (student, parent, school staff (zip code), researchers, health professionals, policy analyst) by
Zip Code
Other demographics to be entered later (ie. type of family, income etc.)

Behavior patterns like in your description reduce humans to a simplified machine that follow a restricted stimulus-response model. But humans have unique skills and spontaneous insights that can't be measured or predicted. Also, observations are subjective, not objective, and may be erroneous. Your database model ignores that. So it will be a kind of "rubbish in rubbish out" as we say here in Germany.

Beside that, I urge to preserve privacy rights of the students. It should be checked if the database conforms to the local privacy laws.

As long as there are guns allowed in the US, and especially in schools, you will not change much just with a database, or with Elasticsearch.

Anyway, I wish you good luck. Don't be frustrated if the project does not turn out all right. Violent-free schools are a valuable goal, I wish I had such a school in my youth.

Thank you for your comments. When we stop
trying to help reduce the 2 million cases of school violence each year in American schools Maybe I have a minute to discuss the phenomenon of ideological, solipistic and tautological projection.