AI for Qualitative Impact Evaluation

Share on facebook
Facebook
Share on google
Google+
Share on twitter
Twitter
Share on linkedin
LinkedIn

How Artificial Intelligence can be used to analyse Qualitative Data for Impact Evaluation

Natural Language Processing (NLP) is a sub-field of artificial intelligence that allows computers to process, understand and interpret natural human languages such as text and speech. Current approaches to NLP are based on machine learning (ML)  such as examining patterns in natural language data, and using these patterns to improve a computer program’s language comprehension. 

Impact Evaluations are mostly based on quantitative data. Questionnaires, institutional data end so on, are used to statistically estimate the effect of a given program.

On the other hand there are aspects that quantitative data cannot capture. Qualitative data can provide deep insights into how programs or policies work or fail to work and more compelling accounts of success and failure. It can be challenging, however, to know what to do with the data generated from qualitative approaches for a number of reasons:

-volume of material

-numerous possible ways of analyzing it

-overwhelming

-difficult to reasonably disseminate findings end results.

In this perspective, Natural Language Processing becomes a valuable tool to automatically extract clear information from a huge amount of text end highlights key aspects of the effects of the program as reported by the participants. NLP can be used to make the most out of text data and to automatically gather the following set of valuable information/outcomes.

  • Categorising content
    This means that the major keywords of the text are identified in order to group documents by similarity .
  • Writing a summary of the text
    Make a short summary from a text of a bigger volume.
  • Sentimental analysis
    Identify what kind of mood the author or speaker was in when producing the text, or what their general opinion is. This can be determined from vocabulary and grammatical structures used in the text.
  • Syntactic analysis
    By analysing syntactic structures and words of the text together can help determine when and where the text was produced.
  • Semantic analysis
    Semantic analysis allows to understand the interpretation and meaning of words and the way sentences are structured.

More to explorer

Landfill Monitoring

The purpose of this project is to define and develop an effective and capable monitoring system that can count trucks, estimate their load and measure the daily quantity of waste that goes to landfill and summarize the materials that get recycled.

Soil Fertility Monitoring in an Agroforestry System

How soil is impacted by different practices? How dose it change over time? These are few of the questions that we are trying to answer in an experimental project by collecting detail data on many soil properties with the use of state of the art sensors and data analysis.

Add a Comment

Your email address will not be published. Required fields are marked *

Need more information about our services?