Using LLMs to Extract Data from Text
Using LLMs to Extract Data from Text| CRIME De-Coder
I detail how to evaluate the cost and benefits of gun shot detection, from lives saved to reductions in crime. Only areas with very high levels of gun violence do I think it is currently cost effective due to lives saved. There is not consistent evidence it reduces crime.| Crime De-Coder
Dashboard to show how much students spend on books at GSU. Tech stack: Cloudflare R2 to host parquet files, DuckDB WASM to query the data client side, and D3.js to create a sortable table.| Crime De-Coder
Denver's STAR Program and Disorder Crime Reductions| CRIME De-Coder
Analysts should learn SQL to get ad-hoc queries faster, make your work more reliable, and have more skills that are transferable to other analyst positions.| Crime De-Coder
Post 005 Alt Academic newsletter. Up and coming tech cities, increase font size in plots, happy holidays!| Crime De-Coder
What distinguishes CRIME De-Coders training services from other providers is technical expertise, as well as building custom courses to help your crime analysis unit on specific projects.| Crime De-Coder
Post 004 Alt Academic newsletter. Apply a bunch advice, civil service jobs, don't change directory in scripts.| Crime De-Coder
Post 003 Alt Academic newsletter. Healthcare jobs and intro to machine learning with Statquest.| Crime De-Coder
Post 002 Alt Academic newsletter, intel jobs and creating in memory database in python to test SQL| Crime De-Coder
Post 001 Alt Academic newsletter| Crime De-Coder
Alt Academic Newsletter, in which I discuss alt-academic jobs and give professional development advice for PhDs looking into non-traditional professor roles.| Crime De-Coder
Even though GenAI is all the rage, I discuss more common applications of natural language processing (NLP) in crime analysis in this post: named entity recognition, semantic similarity search, and supervised learning with text inputs.| Crime De-Coder
I have a highly advanced network of individuals with PhDs looking for advanced roles. This post discusses why many professors are good candidates for tech jobs, and how I can recruit them for your agency.| Crime De-Coder
It is becoming more common for police departments across the US to publicly share crime data. This post discusses the types of data PDs share, and the benefits to doing so.| Crime De-Coder
Fully automated dashboard with up to date Dallas crime trends. Tech stack: serverless dashboard using pyscript and automated daily refresh using github actions.| Crime De-Coder
Establishing foreseeability in premises liability cases is easy. Here I describe more empirically rigorous ways to establish whether a premise has (or has not) provided reasonable security.| Crime De-Coder
Social Network Analysis (SNA) is an important skill for crime analysts to learn. This post is a primer on what SNA is, where SNA data comes from, and example use cases in crime analysis.| Crime De-Coder
Demo tool to estimate the weighted displacement difference test to estimate crime count reductions compared to control areas.| Crime De-Coder
Demo tool to upload your own social network and identify prioritized members for a focused deterrence intervention.| Crime De-Coder
Converting stat. code to open source (e.g. SAS to python), building data visualization websites, and scientific writing are three services CRIME De-Coder has offered past clients across a wide set of domains.| Crime De-Coder
Dashboards that are months old are not useful, they should be automated to include recent data.| Crime De-Coder
Percent change is not a good metric with crime data and encourages chasing the noise. Use an alternative simple statistic CRIME De-Coder developed, the Poisson Z-score, instead.| Crime De-Coder
Baltimore has a hotspot of crime that costs the police department over a million dollars a year.| Crime De-Coder
CRIME De-Coder's services are distinguished via quality, providing computer code to be able to audit and replicate results, and cost.| Crime De-Coder
Examples of program analysis services CRIME De-Coder can conduct.| Crime De-Coder
Examples of predictive analytic services CRIME De-Coder can conduct.| Crime De-Coder
Examples of process automation services CRIME De-Coder can conduct.| Crime De-Coder
Civil Litigation | Statistical Consulting | Racial Bias | Cost of Crime| CRIME De-Coder
AltAc Newsletter, advice about leveling up and volatility in the market| CRIME De-Coder
BOLOs have less value than these other types of products crime analysts can make| CRIME De-Coder
Post 022 Alt Academic newsletter. Can apply more than once, using python to manipulate mouse for scraping.| Crime De-Coder
Research on detecting lying in 911 calls is junk, and I review a recent paper that makes a major mistake that greatly inflates its accuracy.| Crime De-Coder
Post 021 Alt Academic newsletter. Don't worry about stack ranking, consulting travel jobs.| Crime De-Coder
Post 020 Alt Academic newsletter. Rejection is not a normal part of other jobs, project manager if you are not quant, scraping via APIs or playwright.| Crime De-Coder
Example police attitudes survey. Using PHP and Google Sheets as a backend.| Crime De-Coder
Instead of relying on grants to outsource research, it makes sense for private and public sector agencies to contract directly with researchers. It is much faster and better savings for the researchers themselves.| Crime De-Coder
Post 019 Alt Academic newsletter. Practice your talks, startup job focused on surveys, Willison blog to keep up on LLMs.| Crime De-Coder
Many analysts struggle to connect to data that is crucial for their job. In this post I go over my approach to figure out the proper ODBC connections to query record management systems using python.| Crime De-Coder
Artificial intelligence has a ton of buzz, but not much substance in current public discussions. This post goes over some of the popular AI applications in policing, gives my opinion of their feasibility, and discusses some examples of AI that are over promising what they can accomplish.| Crime De-Coder
Post 018 Alt Academic newsletter. Social media experience, UX researcher, Research position and Chase, and use Quarto.| Crime De-Coder
Updated demo for interactive map. Built using python + folium. Incorporates custom looks and interactions.| Crime De-Coder
Post 017 Alt Academic newsletter. Salary ranges can be misleading, creating LLC, updated jobboard link.| Crime De-Coder
python | crime analysis | temporal analysis | data visualization| CRIME De-Coder
How Automated License Plate Readers work and implications for privacy| CRIME De-Coder
Post 016 Alt Academic newsletter. Python self-published book is out! For those looking to get started in python this is my recommended resource.| Crime De-Coder
Data Science | Crime Analysis | python | book | epub | SQL | pandas| CRIME De-Coder
Slides for webinar hosted by the Carolina Crime Analysis Association| Crime De-Coder
Post 015 Alt Academic newsletter. State researcher job positions, PhD can be applied to many fields, probabilities and expected values.| Crime De-Coder
Response to NIJ's request for input on Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence. I provide a working definition of AI, suggest data transparency is key to for algorithms to be trustworthy, and that NIJ should expand funding for more original development of AI models.| Crime De-Coder
Poisson | Crime Trends | Crime Analysis | Excel | R | python| CRIME De-Coder
Post 014 Alt Academic newsletter. Sending emails are worth it and intro to docker.| Crime De-Coder
geocoding | python | ArcPro | ESRI| CRIME De-Coder
Demonstration interactive hot spot map of traffic accidents in Austin. Tech stack: leaflet and python/folium to create customized looks and interactions.| Crime De-Coder
Post 013 Alt Academic newsletter. All about how CICD is an important quality control mechanism for your code, unit testing code, and github actions.| Crime De-Coder
To better meet police departments needs, CRIME De-Coder builds| Crime De-Coder
AltAc Newsletter | machine learning | tech jobs| CRIME De-Coder
This blogpost shows how to use a comparison area to determine if your CompStat intervention has been successful in reducing crime.| Crime De-Coder
AltAc Newsletter | python | R | SQL | project structure| CRIME De-Coder
This post discusses the two major types of surveys police departments conduct; post-contact and general attitude surveys. It also discusses how CRIME De-Coder has developed lower cost methods than many firms that conduct surveys.| Crime De-Coder
AltAc Newsletter | secrets | big data | audio| CRIME De-Coder
This post discusses my vision for the practical application of artificial intelligence regulation in policing; focusing on harms, front end documentation on the validity of the system, and regular monitoring of the system in practice. As well as whom should do the regulation.| Crime De-Coder
This is my presentation for the 2023 FBI Dash conference. Paper on how random forests have greater predictive accuracy than RTM, and how you can use interpretable model summaries to describe hotspots and criminological patterns in the data.| Crime De-Coder
This is my presentation for the Carolina Crime Analysis conference in 2024. Illustrates the use of the WDD stat to evaluate pre/post crime changes relative to control areas.| Crime De-Coder
Post 009 Alt Academic newsletter. My notes on LLMs; deployment hassles, different prompt ideas (RAG, zero-shot, one-shot).| Crime De-Coder
Presentations | javascript | machine-learning | statistics| CRIME De-Coder
This post give advice about how to conceptualize what you place on your dashboard, the types of comparisons you make, and whether alternatives such as auto-alerts or standardized reports are better solutions for your use case.| Crime De-Coder
AltAc Newsletter | portfolio | javascript | php | html| CRIME De-Coder
Post 007 Alt Academic newsletter. People embellishing tech skills vs imposter syndrome, fraud jobs at TikTok and Lexis Nexis, github intro.| Crime De-Coder
Early Intervention Systems (EIS), are algorithms used to identify problematic behavior by police officers. This post goes over what they hare, how they should be used, and gives a price range that CRIME De-Coder can build custom software for your department.| Crime De-Coder
Post 006 Alt Academic newsletter. Banking jobs, CTEs in SQL, product/project managers| Crime De-Coder
Sworn Officer Rates over Time| CRIME De-Coder
Predictive Policing | Forecasting | Early Intervention Systems| CRIME De-Coder
Data Science | Crime Analysis | python | book| CRIME De-Coder
Crime Analysis | Redistricting | Police Performance | Police Workload | Hotspots| CRIME De-Coder
