Browse Prior Art Database

Method to enhance crop yield by means of artificial intelligence and natural language processing. Disclosure Number: IPCOM000234020D
Publication Date: 2014-Jan-07
Document File: 3 page(s) / 32K

Publishing Venue

The Prior Art Database


Enhancing crop yield by means of artificial intelligence and natural language processing. Analytics with application of artificial intelligence and natural language processing can be used effectively to enhance the yield a given crop in a demographic area. The system described below explains a learning model based on inputs from multiple structured and unstructured data points such as weather pattern, genomics, soil, GIS, user information, chemical information of soil and fertilizer etc. The main advantage of such a system is that a farmer gets to know the best known methods or new innovative methods to improve the yield of the crop tailored to his personal choice of crops.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 48% of the total text.

Page 01 of 3

Method to enhance crop yield by means of artificial intelligence and natural language processing.

One of the applications is in the field of crop yield and plantation advice. Large amount of information exists on these topics in different forms. There exist huge databases of literature on soil type, weather conditions, and weather patterns. For example, the data may come from books, journal articles about

yield etc, and web forums.

Additionally, there is a growing use of location specific sensors that act as a data point, giving real time information about the weather conditions. Other sources of the data include Geographical Information Systems (GIS) information on soil texture and so on. Web data about the region, the soil type and quality found in the region, weather patterns/ articles about yield and so on.

There are companies such as ESRI*, in the location analytics business. These companies provide high-quality data for GIS analytics. Example, one can get imagery data, base maps, and climate change data. ESRI - provides analytics and data based on GIS information but it doesn't go far enough with combining a natural language processing system with this analytics data to provide key insights for agriculture.

The main idea of this article is to propose an IBM Watson* solution that gives recommendation to enhance the yield of a crop. Given a location and some simple information such as available crop types and current farming practice being followed, a user in a remote part of the world can ask on best practices on crop rotation, type of crop and other methods in agriculture for the best possible yield. Watson would be able to give them a reasonable answer, with confidence and evidence to support the case.

Learning Model

This article proposes a learning model based on the key inputs from soil chemistry, fertilizers, pesticides, crop rotation, crop properties and weather patterns, shifts in weather and harvest time. The learning model also considers the positive and negative impacts due to various factors such as over use of pesticides and fertilizers, soil salinity etc.

The core idea of this article is to cross mine and gain insight from various data sources, such as:

(a) Natural language agricultural articles.

Example: articles on organic quality of the soil, data on crop rotation etc.

(b) Weather pattern data

(c) Sensor based data

(d) Various databases for crops, fertilizers and pesticides.

(e) User interaction and sensor based data on soil structure


This would find a major application in the agricultural industry. The farmers can readily get crucial information on do's and don'ts for improving the crop yield from software as a service system. Example, the system can recommend improvements/adjustments to the farming method that will not only help improve yield per hectare but may also help maintain the ecological balance of their lands for future harvest.


User Query: The leaves of my papaya trees ar...