0

Context-Driven Natural Language - DZone AI

 5 months ago
source link: https://dzone.com/articles/get-context-driven-natural-language-processing-for
Go to the source link to view the article. You can view the picture content, updated content and better typesetting reading experience. If the link is broken, please click the button below to view the snapshot at that time.

Get Context-Driven Natural Language Processing for the Best Search Results

Augmented Analytics that includes context-driven Natural Language Processing (NLP) will provide the kind of support your business needs to achieve the results.

by

·

Jun. 06, 22 · AI Zone · Opinion

If your business is focused on data-driven, fact-based decisions, your business users may be leveraging an analytics solution to gather, find and analyze data. Business goals include improving results and productivity and getting the best results out of your data, as well as gaining meaningful insight into data. But, you certainly want to accomplish all those goals without frustrating business users or forcing them to adopt tools that do not add value to their day-to-day workflow and tasks. 

When a business sets goals and establishes metrics to determine the value of an analytical solution and a business user analytics initiative within the enterprise, the management team often fails to focus on the more subtle but powerful concept of efficacy. Just how effective is the search process and, equally important, the results produced by that search. I suppose a user can quickly and easily ask a question without having to find and choose columns or fields. In that case, they can avoid the time and complexity of putting together an effective search and the frustration of receiving results that do not fit their requirements. 

When a business envisions a data-driven environment and establishes the foundation with an analytical solution, it should expect user adoption and a more data literate team environment. But, without effective search tools, that vision will fall short.  

Augmented Analytics that includes context-driven Natural Language Processing (NLP) will provide the kind of support your business needs to achieve the results you envision.

Context-driven searches using Natural Language Processing (NLP) allow users to go beyond restrictive searching based on column filters to provide context such as season, time series or range, polarity, and they address abbreviations and analyze phonetics to handle misspellings, etc. This approach allows users to ask a question using a more human, conversational approach without having to consider which columns or fields they need or how to ask the question to get an answer that provides meaningful insight into their question without frustrating the user by returning no results or results that have nothing to do with the question. 

‘If a user can quickly and easily ask a question without having to find and choose columns or fields, they can avoid the time and complexity of putting together an effective search.’

In order to understand the very real value of this type of tool, it is helpful to consider a few examples of business applications.

Context: Time Series

  • Benefits

Enter a question and receive results based on absolute time or a range or relative peri.  

  • Sample Date Range Question

Which sales representative sold the most pancake mix from April 2015 to May 2019?   

  • Sample Relative Time Period Question

Who sold the most cake in Phoenix, Arizona, in the last quarter of 2019?  

  • Sample Absolute Time Question

What is Bill Jones’ best-selling product for this year?   

Context: Synonyms, Phonetics & Abbreviations

  • Benefits

Enter a question, and the system will recognize and process information, correcting for spelling errors, abbreviations, and related words. 

  • Sample Phonetics-Based Question

Who sold the most fruit juice in Phoenix, Arizona, on Christmas of 2018?

  • Sample Abbreviation Question

Who sold the most fruit juice in Phoenix, Az, on Christmas of 2018? 

Context: Aggregation

  • Benefits

Enter a question to understand results for averages, minimum, maximum, first, last, sum, counts, etc. 

  • Sample ‘Average’ Question

What is the average list price for ginger tea in the Western Region? 

  • Sample ‘Count’ Question

Tell me the number of sales representatives in Arizona.

Think about context-driven searching as the foundation for more collaboration and improved insight. If you are sitting in a staff meeting or at a conference and listening to a presentation, the presenter will often provide time at the end of the session for questions. After considering the topics and issues presented, you may need answers to further clarify the information you are processing and make the best use of the information presented. Context-searching provides the same kind of support. 

If a sales manager is considering how to optimize her sales team and leverage customer product and purchasing preferences, the business would want to encourage creativity and out-of-the-box thinking. With context-driven searching, the sales manager can ask all kinds of questions to test theories and hypotheses. She can explore the team results and discover interesting and useful insight into sales, regions, team vs. individual results, and more – all without having to understand data extraction techniques or having to create complex searches using column filters and other data structural components, or navigating menus or scripts to elicit answers from the data. What you want (and what you will get) from NLP searches is a ‘what you see is what you get approach to searching.   Think of a question and type that question, and you’ll get results.  

When a business is choosing analytical tools, it is wise to choose tools that support the way humans think and communicate and provide results in a clear fashion to support fact-based decisions. Keep it as simple as possible. Ask a question and get an answer. That is the most effective way to search for information. 


About Joyk


Aggregate valuable and interesting links.
Joyk means Joy of geeK