Skip to main content
  1. Hem >
  2. Om Fujitsu >
  3. Case Studies>
  4. Data Analysis System reduces efforts for web investigation at Japan National Police Agency

Data Analysis System reduces efforts for web investigation at Japan National Police Agency

A data analysis system combining external information from websites and criminal investigation information from within the Agency

The challenge

Information regarding crime can be found within the Internet, but has been difficult to utilize due to the massive amount of data available. Detecting only the information relevant to the crime and combining it with criminal information stored at the Agency for analysis required expert know-how and significant time and effort from the analysts.

The solution

  • An automatic search crawler which selects/collects only the URLs intended by the searcher/analyst
  • Significantly decrease analysis time with parallel execution of existing analysis applications on Hadoop
  • Analyze natural language in collected information from the Internet and the Agency’s internal database, adding indexes useful for research

The benefit

  • Automatically searches through the massive amount of information on the Internet for things necessary in the investigation, reducing the number of searching hours significantly (narrowed from 1.2 million to 279 hits)
  • Scale-out type architecture combining parallel distributed processing and machine learning reduces analysis time from 3 days to 6 hours
  • Machine learning and automatic tagging can be applied to information from the Internet and the Agency, resulting in sharing and utilization of experienced analysts’ analytical viewpoint know-how
  • Based on a general package, the system can be implemented in a short time. (In only 3 months: From development to delivery, beginning of service)

Products and Services

  • ODMA Website Analysis PKG

Ladda ner

Read the full National Police Agency case study (993 KB/A4, 1 page)