Emergency calls, audio/video from body-worn cameras and interview rooms, and other critical communications can hold crucial information for expediting investigations or solving crimes. Yet, it often takes hours to find the relevant audio recordings. Not anymore!
NICE Inform Audio Analytics, a new analytics solution from NICE Systems, accelerates and simplifies this process, enabling precise, fast searching and categorization of audio content using key words and phrases.
Currently PSAPs and security operations that record audio for investigations, quality assurance, or compliance are constrained to using some combination of simple metadata, such as call date and time, call taker position, and the calling party’s phone number and location (ANI/ALI) to find a recording of interest. If they’re able to locate the call they’re looking for, it’s then necessary to listen to the entire voice recording to determine what was or wasn’t said.
NICE Inform Audio Analytics offers significant improvements by enabling the search of audio content based on words and phrases spoken within the recording. Audio recordings are automatically analyzed right after they are recorded, or in the case of body-worn cameras or other off-line audio, the audio analytics can be applied when audio is uploaded to the system.
As identified words are indexed, audio recordings are then assigned to one or more user-specified categories based on the words or phrases found. This helps in quickly pinpointing calls relevant to a specific investigation, or in identifying quality assurance or compliance issues.
“The proliferation of big data in policing is a major trend and significant challenge,” said Chris Wooten, Executive Vice President, NICE Security Group. “With more and more digital media to analyze, now including video and audio from body-worn cameras, public safety and security operations will increasingly turn to analytics to gain valuable situational awareness and identify critical evidence. With the addition of Inform Audio Analytics, NICE’s impressive portfolio of audio and video analytics solutions are set to propel policing into the 21st century by helping organizations leverage digital data to its fullest potential, while enhancing efficiency and operational intelligence.”
How it works:
Under the hood, NICE Inform Audio Analytics uses N-Best indexing, which simultaneously matches multiple words or phrases to the same audio along with an estimated confidence level for each word – for example, ‘gun’ at 80% or ‘gone’ at 70%.
This greatly improves the probability that all occurrences of a target word or phrase will be found, even when audio quality is degraded by background noise. A powerful index maps the words and phrases to the calls that contain them. Searches on audio indexed in this manner return results in seconds.
More conventional analytics, based on phonetic indexing technology, result in significantly slower searches because the solution has to build the phonemes into words and phrases every time a search is run. To further enable rapid analysis, NICE Inform Audio Analytics plays back from the precise location of the word or phrase, so it’s not necessary to listen to the entire call to find the relevant audio section.
NICE Audio Analytics can be used for:
•Investigations – To retrieve information faster, or find calls that are part of a build-up to an incident that can yield insight into the cause or reveal witnesses.
•Compliance – To ensure all protocols are being followed according to specific compliance requirements and identify non-compliant employees for remediation.
•Quality assurance – To improve efficiency and effectiveness of QA by isolating specific types of calls for targeted QA review; to identify problem calls where inappropriate language was used; to automate call analysis to ensure that procedures are being followed and employees are consistently performing to the highest standards; and to identify gaps in training or policies that need to be addressed.
•Operational intelligence – To automate trend reports that highlight the types and quantities of calls taken over specific time periods.
