 This is part 2 of my research demo. Here is our research command. I have a list here that contains URLs. So as you know, with our research name URL, I can extract these URLs from this list. So now again I'm going to look for domain names. So simple domain names that are a sequence of letters. And then with .com. So here I have domain names and you can see DJ Stavins, Google, and Microsoft.com. And then two names that appear to be just random letters. Now I have also built into our research features that allow you to select names, select strings, based on the fact if they look like English words or not. So you can actually filter out gibberish like these random words here. And I do this with a module that I developed and is based on an open source module. What I do is I do frequency analysis of two letters, of two following letters. And if the frequencies of those letters are much frequencies that you find in English text, then I assign it a high score. And if they don't, I assign it a low score. And that's how I select. Now how do you use this? Let me first show you a regular expression command. So if you type this, then you have actually added a command to your regular expression. So this here is a command and there's no influence whatsoever on regular expression. Now what I did is the following. If you start regular expression with a command and the command is extra equal and then some instructions, then you are able to invoke the special functions like gibberish filtering here. So what I'm going to do here is say I'm going to use the S feature. S feature stands for sensical. It allows you to select or deselect sensical text. So not gibberish. And here I'm going to use the option S for sensical. So you see here for my list now I select all domains that look like English words. If I say gibberish, T, I do the opposite. I select domains that look like random that don't look English. So that is one feature here, sensical detection where you can select sensical or gibberish text. I also added white list and black list. So let me go back here. Here I select domains. Again I can add a command like this, extra equal. And now I'm going to say exclude from blacklist. You can see here everything is selected except google.com. That is because google.com is in the blacklist. So I can now select this on regular expression and I can also use blacklist to remove some of those selections. I also have a white list, that's me. And if I want to use that white list, so I include now, I don't exclude, I include whitelist.pxt. And here I'm selected. It is possible to combine features. So let's go back here and select sensical words like this only to three domains that look like English words. And then I'm going to exclude from google via the blacklist. So exclude blacklist.com. Sorry. Like this, blacklist.com. No, sorry, it's not blacklist.com. It's blacklist.pxt. Here you see google has been excluded. So you can combine features like sensical here and whitelisting, blacklisting.