The purpose of your search is simple: getting what you want. Despite this simplicity, the process can be very complicated.
First of all, when we search, we usually look for words that are most familiar to us. For example, when we want information about “killer”, we do not immediately take into consideration other keywords like “murderer”. Search engines and search tools may not be intelligent enough to look for those semantically related words for us. You may find yourself visiting page after page, jumping from one word to another, or changing your search term over and over to find what you really want.
Furthermore, even after performing a Google© search and finding what looked like good results, you click a link and find yourself still buried inside a mountain of text, with a dense page full of information. You wanted a quick answer, but ended up looking through long webpages with no idea where the information you want is.
You might scroll through the page to look for the information, maybe you hit Ctrl+F and perform a Find-On-Page search, or maybe you click the back button and continue on to another webpage – either way, you’ve lost time.
What if Ctrl+F can understand what you want? What if the browser find function can highlight paragraphs that only contain the information you want?
Twinword Finder searches a web page for sections containing words that match or words that are related to your search terms. It helps save time by jumping straight to what you want, to the relevant places that Ctrl+F might have missed and forced you to find yourself.
With this browser extension, you can also control the search distance to get more closely related words or more loosely related ones.
Get more specific by entering multiple search terms to match only words connected to all of them. The more you enter, the more specific your results are.
Now available for Google Chrome browser.
I am interested in natural language processing that combines computer science and linguistics. I believe the current level of AI and NLP technology is not enough to meet people’s expectation due to lack of qualitative training data for machine. I’d like to overcome this using human computation and data collection platform.