Lexalytics Joins InMoment | Accelerating Global Delivery of Structured and Unstructured Text

Semantria for Excel / Quickstart

Setup

1 First, if you haven't already, activate your Semantria account

During signup we sent you a confirmation email (check your Spam and Junk folders if it's not showing up)

If you're already a confirmed Semantria user, you can proceed to the second step running an analysis.

2 Download Semantria for Excel

System requirements:
1) Microsoft Windows XP, Vista, 7, 8, 10
2) Microsoft Excel 2010, 2013 or 2016 (installed desktop version - trial or online-only versions will NOT work)

Mac Users: We have no native support for Mac, but you can run it through a Windows virtual machine setup

System recommendations:
1) 64bit Microsoft Excel
2) >4GB RAM
3) Dual core or better CPU

Is my Excel 32 or 64 bit?

Note: Check this as it is possible to have 32 bit Excel running on 64 bit Windows

Excel 2010 File > Help > About Microsoft Excel
Excel 2013 & 2016 File > Account > About Excel

 32-bit Office  64-bit Office

3 Run the setup file on your computer

  1. Close Microsoft Excel if it is running
  2. Double-click on the installation file (Semantria.Excel.Setup.xXX.exe) and follow the on-screen instructions
  3. Complete the whole setup process

4 Enter your credentials in Excel

  1. Open Microsoft Excel
  2. A Semantria sign-in window will open
  3. Enter the username and password you provided during the signup process. You can also enter them after setup under [Settings > Sign-in] in the Lexalytics ribbon tab.

Running Your First Analysis

  1. In the Lexalytics tab in Microsoft Excel, click on Start to open the New Analysis wizard. (Troubleshooting)
    Start button
  2. Import your text to analyze. If you want, use our sample data set below.

    Bellagio Reviews Dataset (.xlsx file)

    Analysis wizard step 1 Select Data Source
  3. Categorize your data by ID, metadata, and the text to analyze. (If you don't see column names, click on "First row has column headings")
    Analysis wizard step 2 Categorize Data
    Analysis wizard step 3 Categorize Data
  4. Select the rows to analyze.
    Analysis wizard step 6 Rows to Analyze
  5. Name your project, select the appropriate language and configuration, in this case English, and click Next.
    Analysis wizard step 7 Analysis Settings
  6. Select the desired reports under Summary Reports and Detail Reports and they will generate. Clicking on the top half of the report buttons will auto-generate all of the reports in that category. Clicking on the bottom of the button will allow you to select individual reports. Then you may want to click "Analytics panel" up next to the Start button in order to review the reports at full width.
    Quickstart analysis results

You've completed your first analysis! For more help see our troubleshooting, step-by-step tutorials, customization tips, and fine-tuning

Reports

Summary Reports*

Except for the Query Co-occurrence report, all Summary reports will give you the top items of whatever the report type is. This report shows how many of those items were Positive, Neutral or Negative, as well as the total number of occurrences for that item. Summary reports also contain two Excel built charts based on the content of the report. These visualizations are very basic. If you are interested in more visualization tools Semantria Storage and Visualization (SSV) might interest you.

The Query Co-occurrence report is the only unique Summary report. This report shows the how many documents hit on a cross section of the queries in the configuration.

  1. Sentiment Phrases
  2. Themes
  3. Entities
  4. Queries
  5. Concept Topics
  6. Autocategories
  7. Query Co-Occurence

Detail Reports*

Detail reports contain all the detailed output that Semantria generates when analyzing content. The Document Overview detail report will be generated at the same time as any other detail report, as detail reports have links in the ID column back to the Document Overview report.

  1. Document Overview
    Document ID:
    The ID of the document
    Status:
    The status returned from Semantria. “PROCESSED” means that the Document was processed correctly. Anything else will indicate a reason as to why the Document was not analyzed.
    Source Text:
    The text of the document that was analyzed
    Summary:
    A summary of the Document, the length of which depends on the summary length setting in the configuration that was used to analyze the content
    Detected Language:
    The language that Semantria believes the Document to be written in. NOTE this is NOT the language that the content was analyzed in. That is determined by the language of the configuration that was used to analyze the content.
    Detected Language Score:
    A score of how confident Semantria is that the Detected Language is correct
    Document Sentiment:
    The numerical sentiment score assigned to the Document
    Document Sentiment:
    The polarity of the sentiment score (Positive/Neutral/Negative)
    Metadata:
    If the user attached any Metadata to the analysis then those columns will be displayed after the Semantria output
  2. Words
    Word:
    A Word
    Type:
    The type of Word
    Number of Mentions:
    he number of times that Word occurs in all the documents
    Documents Count:
    The number of documents that Word occurs in
  3. Sentiment Phrases
    Document ID:
    The ID of the document (This is a link back to the Document Overview report)
    Highlighted Text:
    If the configuration used to analyze the content has Mentions enabled, then the highlighted phrase will appear in context here.
    Phrase:
    The sentiment phrase
    Phrase Sentiment:
    The numerical sentiment score assigned to the sentiment phrase
    Phrase Sentiment +/- :
    The polarity of the sentiment score (Positive/Neutral/Negative)
    Phrase Intensifiers:
    If the phrase is being intensified, then the intensifier will be listed here. Eg. "very" or "more" or "super" etc.
    Phrase Negators:
    If the phrase is being negated, then the negator will appear here. Eg. "not" or "no"
    Metadata
    If the user attached any Metadata to the analysis then those columns will be displayed after the Semantria output
  4. Themes
    Document ID:
    The ID of the document (This is a link back to the Document Overview report)
    Highlighted Text:
    If the configuration used to analyze the content has Mentions enabled, then the highlighted theme will appear in context here.
    Theme:
    The detected theme. NOTE Themes are autodetected and not configurable
    Strength:
    Relevancy of the theme
    Theme Sentiment:
    The numerical sentiment score assigned to the theme
    Theme Sentiment +/- :
    The polarity of the sentiment score (Positive/Neutral/Negative)
    Theme Sentiment Evidence:
    Amount of sentiment evidence for this theme
    Theme Stemmed Form:
    Stemmed version of the theme
    Theme Normalized Name:
    Normalized version of theme
    Metadata:
    If the user attached any Metadata to the analysis then those columns will be displayed after the Semantria output
  5. Entity Themes
    Document ID
    The ID of the document (This is a link back to the Document Overview report)
    Highlighted Text
    If the configuration used to analyze the content has Mentions enabled, then the highlighted entity theme will appear in context here
    Entity
    The entity
    Entity Type
    The entity type
    Entity Theme
    The entity theme
    Entity Theme Sentiment
    The numerical sentiment score assigned to the entity theme
    Entity Theme Sentiment +/-
    The polarity of the sentiment score (Positive/Neutral/Negative)
    Entity Theme Sentiment Evidence
    Amount of sentiment evidence for this theme
    Metadata:
    If the user attached any Metadata to the analysis then those columns will be displayed after the Semantria output
  6. Entities
    Document ID:
    The ID of the document (This is a link back to the Document Overview report)
    Highlighted Text:
    If the configuration used to analyze the content has Mentions enabled, then the highlighted entity will appear in context here.
    Entity:
    The entity
    Entity Type:
    The entity type
    User-Defined Entity:
    A “yes” here indicates that the entity was defined by the user. A “no” indicates that the entity was autodetected.
    Entity Sentiment:
    The numerical sentiment score assigned to the Entity
    Entity Sentiment +/- :
    The polarity of the sentiment score (Positive/Neutral/Negative)
    Entity Sentiment Evidence:
    Amount of sentiment evidence for this entity
  7. Queries
  8. Document ID:
    The ID of the document (This is a link back to the Document Overview report)
    Highlighted Text:
    If the configuration used to analyze the content has Mentions enabled, then the highlighted query keyword will appear in context here.
    Query Category:
    The query that was hit on
    Query Category Sentiment:
    The numerical sentiment score assigned to the Query
    Query Category Sentiment +/- :
    The polarity of the sentiment score (Positive/Neutral/Negative)
    Query Category Relevancy:
    The number of query terms that hit in the document
    Metadata:
    If the user attached any Metadata to the analysis then those columns will be displayed after the Semantria output
  9. Concept Topics (Sometimes referred to as User Categories)
    Document ID:
    The ID of the document (This is a link back to the Document Overview report)
    Source Text:
    The text of the document where the Concept Topic was found
    Concept Topic:
    The Concept Topic
    Concept Topic Sentiment:
    The numerical sentiment score assigned to the Concept Topic
    Concept Topic Sentiment +/- :
    The polarity of the sentiment score (Positive/Neutral/Negative)
    Concept Topic Strength:
    The level of confidence that Semantria has that this Concept Topic applies to the document
    Metadata:
    If the user attached any Metadata to the analysis then those columns will be displayed after the Semantria output
  10. Autocategories
    Document ID:
    The ID of the document (This is a link back to the Document Overview report)
    Source Text:
    The text of the document where the Autocategory was found
    Autocategory:
    The Autocategory
    Subcategory:
    If there is a Subcategory, it will appear here
    Autocategory Sentiment:
    The numerical sentiment score assigned to the Autocategory
    Autocategory Sentiment +/- :
    The polarity of the sentiment score (Positive/Neutral/Negative)
    Autocategory Strength:
    This is the relevance score for the Autocategory
  11. Intentions
    Document ID:
    The ID of the document (This is a link back to the Document Overview report)
    Source Text:
    The text of the document where the Intention was found
    Intention Type:
    The type of Intention
    Who:
    Who does the Intention belong to
    What:
    What does the intention refer to
    Evidence:
    Evidence of the intention
    Metadata:
    If the user attached any Metadata to the analysis then those columns will be displayed after the Semantria output
  12. Machine Learning Models: This report is only available to those that have machine learning models installed in their configurations. This is not a standard feature, if you are interested in learning more about machine models please contact us.

Tuning Reports*

  1. Uncategorized Documents: This report lists the documents that did not hit on any queries, and as such can be used to tune queries.
  2. Possible Sentiment Phrases: This report lists bi- and trigrams that could possibly be sentiment phrases. Note that these are not actually sentiment phrases, but rather they fit the pattern that other sentiment phrases do, so they are listed here as possible sentiment phrases that a user could add to their configuration.
  3. Query Comparison: This report compares the number of query hits from two separate analyses. This is useful in comparing an older analysis to a new one where you have made query changes.
  4. Sentiment Phrases for Queries: This report details the sentiment phrases that give a query its sentiment. The columns in this report are a combination of certain columns from the detail reports for queries and sentiment phrases, refer to the detail report information above for any clarification.
    Note that you can see multiple sentiment phrases for one query result in this report.

* available tabs depend on the licensed features

Troubleshooting

The start button is disabled / greyed out / not able to be clicked
Without closing the original Excel workbook, open a second excel workbook. You should now have two open Excel windows. Return to the original window and check the start button. It should now be enabled. If it is still disabled, try rebooting your computer.
Not requesting documents / not receiving documents
Log out and back in.
API error message
Click OK on the error message, check that you are logged in, then continue with your work as needed.
Waiting for documents for an excessive amount of time
If you sent your documents over 24 hours ago then the results of that analysis are no longer going to be in the queue waiting to be sent to your machine. You will need to reprocess your documents. If this was a very large amount of documents that you sent and did not receive results for, then please contact support to discuss reimbursing your credits.
Blank row error message
If Excel hits a blank row in any of the Text columns that you choose than that row will be considered the end of the worksheet. If you want to analyze content that has blank responses or empty text fields, you need to replace those blanks with a placeholder such as “N/A”.
Tab names
The name of the reports will be the name that you specified in the analysis followed by the type of report. Note that the name of the report will supersede the name of an analysis, that is to say if you analysis name is long it will get truncated so that the type of report is clearly displayed.
Text display issues possible on Macs with Retina displays running VMs and Windows machines with high resolution displays.
Typically restarting your computer resolves display issues, but in some rare cases this is not the case. In this event we suggest that you play with the scaling of text in your settings as well as checking to see if your computer is auto-scaling programs.
Semantria for Excel is running very slowly
The Excel plug in stores all results in a local database. As such if you run a lot of content your database might get very large. This can cause slowdowns within Excel. A simple way to work around this is to save your reports as Excel workbooks, and do not rely on saving them in the Analytics Panel. To delete an analysis double click on the analysis in the panel, this will open a menu where you can choose to delete the analysis.
Another issue?
Having another issue that isn’t seen here? Email support@lexalytics.com with a detailed description of your issue and include your log file.
Getting your log file:
  1. Launch the Run application on Windows (Press the Windows button and “r” at the same time for a shortcut.)
  2. In that window enter “%appdata%\Semantria\ExcelAddIn” without the quotes.
  3. Click OK
  4. In the window that opens you should see a “semantria” file of the type “Text Document”. This is your log file and should be attached to any email regarding an issue with the Excel plugin.

Or call us at 1-800-377-8036