Video Tutorials: Semantria Advanced Functionality

  3 m, 35 s

There is no better way to learn something than a video tutorial. I’ve taught myself how to make my own music in a virtual studio, cook delicious meals for friends and family, and fix my toilet so I didn’t need to bother my landlord to call a plumber. We truly do live in the world of DIY (do it yourself).

If you’re new to text analytics and want to understand the basics of how to use this technology with Microsoft Excel, check out the Introduction to Semantria for Excel playlist. 

However, the remainder of this post will cover the advanced features of Semantria for Excel. Wait no longer, here’s a link to the video tutorial playlist: Advanced Functionality of Semantria for Excel

Video 1 – Configurations

Video Tutorial Screencap 1

When you decide you want to start making changes to Semantria, like adding your own custom entities and adjusting sentiment scores for different phrases, you may want to build a new configuration first. This is great because you can play with the new configuration all you want and you won’t mess with the default settings of Semantria.

Video 2 – Sentiment Adjustment

Video Tutorial Screenshot 2

Once you’ve built a configuration, you may want to add some sentiment bearing phrases to the dictionary or adjust some of the default sentiment scores from positive to negative. The classic example comes from a company that manufactures vacuum cleaners. They ran all of their data through Semantria and everything came out negative. The problem was that the word “sucks” bears negative sentiment by default. By adjusting the sentiment score of the word “sucks” from negative to positive Semantria will know that when a vacuum cleaner sucks, it’s actually a good thing!

Video 3 – Customizing Entities

Video Tutorial Screencap 3

You can add entities to any configuration so that Semantria will recognize and classify them according to what “type” of entity they are. The video will also show you how to account for spelling errors or words with multiple spelling variations. For example, the word Brazil can be spelled with an ‘s’ or a ‘z’. You can train Semantria to tag all mentions of both spelling variations, and it will tell you that it’s a country under the column “entity type”.

Video 4 – VLOOKUP

VIdeo Tutorial Screencap 4

There are probably a bunch of other tutorial videos on how to use Microsoft Excel’s VLOOKUP function, but this one is specific to Semantria. Using this allows you to transfer a column of data from one worksheet to another. So if you want to take data from your original worksheet (like name, location, birthdate, etc.) and add it to the Semantria output, check out this video.

Video 5 – Boolean Query Logic

Video Tutorial Screencap 5

If you are unfamiliar with Boolean Operators, this video will go over all of the ones that are used with Semantria. It will allow you to create queries that will quickly classify your dataset with extreme accuracy. The learning curve really isn’t that steep. After writing just a few, you’ll be a pro! Promise 😉

Here’s a list of Semantria’s Boolean Operators:

OR- used to add another term; will broaden search

AND- used to make search more specific; will narrow search

NOT- used to exclude specific term; will narrow search

NEAR/3- used to match terms within (3) characters of each out; can narrow or broaden search

*- used to tag multiple terms that come from the same root word

“ ”- used to tag terms that contain more than one word with a space in between them

Video 6 – Nested Queries

Video Tutorial Screencap 6

The final video in this video tutorial playlist will show you how to use Nested Queries. It’s a relatively new feature in Semantria where you can make references to other query strings, from within the query you’re currently writing. All you have to do is add an open bracket, a wildcard, the name of the query you want to reference, then a closed bracket.

Here’s an example of a nested query:

(*Black) AND (*Cat)

This would tell Semantria to tag documents with black cats, as long as there are already queries made for “Black” and “Cat”.

Black = (black OR dark)

Cat = (kitty OR cat OR feline)

Hopefully this encourages you to watch these videos, and see how all of these features can help you increase the accuracy of your Semantria output.

Categories: Named Entity Extraction, Product Information, Text Analytics