Fuzzy search – Don’t let spelling mistakes become a problem
YOSE has always had an advanced search capability. Even before this release, you could do advanced searches through the YOSE environment such as AND, OR, NOT, just to name a few. So, if you were looking to find John Smith, you could search for “John OR Smith” to get all the Johns and Smiths in just a second.
That’s already powerful, so the question is, how did we make it better?
Well, one thing is that we made it even more accurate by implementing Fuzzy Search. Let’s say, for example, that you are searching for the word cocaine. Someone might be spelling it with a K, and someone else miss-spelled it to cociane.
Now you have three different spellings: cocaine, kokain, and cociane.
Instead of trying different words and spelling, Fuzzy Search can address all those spellings and give you a far more accurate result where kokain and cociane appear in the search result when you search for cocaine.
Let us give another example of a topical situation where Fuzzy Search is extremely useful: Investigating the war in Ukraine.
If you are monitoring the situation in Ukraine as an intelligence or disinformation analyst, you would probably want to search for the names of towns and people. But names in Ukraine are spelled differently depending on where you come from.
That’s why you need something like Fuzzy Search – so you can search through a lot of information at high speed without missing anything important just because it was spelled differently.
Visual grouping is another feature where Ukraine is an excellent example to demonstrate how it works — even though there are many other reasons to use it.
Again, if you are an analyst monitoring events in Ukraine and you have scraped data from Telegram, including a significant volume of images and videos, many of those pictures will likely be duplicates. The reason for that is that telegram groups and channels will re-post and forward content to each other. So, to help sort out those duplicates, we created Visual Grouping.
Visual Grouping is a feature that tidies up your search results and puts duplicates together in a nice group. This will help you investigate, for example, where and when disinformation has been spread in different channels by visualizing all the duplicates in a link chart.
Visual Grouping will not only tidy up your search results but also save you a lot of time that you would have to spend scrolling through duplicates.
In this update, we truly adapted to what’s happening worldwide and made updates to address our users’ issues — duplicates being of them.
Better Face Recognition with even more accuracy than ever before
Facial recognition is one of the most beloved tools by our users. It allows you to find people’s faces in pictures and videos in a few seconds.
Let’s say, for example, you are working with static surveillance. You will probably have hours and hours of video footage to look through – usually from many angles. What most investigators do today is just speed up the footage by 4x.
There are two obvious problems with that.
- It will still take a lot of time.
- The risk of missing something when looking at that speed is significant.
In such a scenario, it might be a person showing up in a window for a split second, which could inevitably lead to the analyst missing vital clues. But if a face shows up for a split second while using YOSE, our face recognition will find it.
By telling YOSE which face you want to search for, it can search through thousands of pictures and videos, group them and present them all to you. So if the person you were looking for was in a video — you will know precisely what second in that video the person appears.
All this was already in YOSE before the new update. What’s new is that we have made it even more accurate and more powerful.
Active thumbnails to speed up your work
The last thing we want to tell you about is Active Thumbnails. This feature will show you if there is an activity in your videos without you having to spend hours looking through them all.
Investigators working with surveillance know how important this is. Because when you sit in front of hundreds of videos, you want to see directly which ones have activity and which don’t. Active Thumbnails will save the investigators a lot of time they would waste looking through all those videos where there is no activity.
These are four new and better features you will have in the latest update of YOSE.
But there is much more! This update is a real game-changer for organizations working with investigations — and we can’t wait to show it to the world!