Many of the things that NLP will achieve are still unknown. It hasn't been advanced enough to be useful in many situations until very recently. Now, there are more applications for this technology than there are people who are capable of implementing it. Here is just a fraction of what you could do, or is already being done:
Natural language processing includes a multitude of complex tasks that have so far prevented the creation of a human-level artificial intelligence. Most software developers won't even know where to begin in solving these problems and you will need a tool that makes it easy for them to utilize the power of a human language understanding software.
Part of speech tagging is the process of assigning a grammatical designation to a word, such as noun or verb. Jemigo uses rule-based algorithms along with a custom-built dictionary to tag each word. This task is complex for various reasons and there is no system yet that has achieved human-level performance.
Words not only have multiple possibilities for parts of speech, but also for meaning. A word can have dozens of meanings depending on the context. Determining the part of speech narrows down the choices and then Jemigo's unique ontology provides a hierarchical and relational data model needed to disambiguate.
It is essential to know when multiple words refer to the same entity in a given text. Pronouns are a great example of this and can sometimes be confusing, even to humans. This feature is still under development, but it will be necessary for a coherent understanding.
NER is the ability of software to identify proper nouns, such as people or locations. This capability is under development.
One of the most challenging and fruitful technologies in the domain of NLP is the extraction of pertinent information from unstructured data. This has endless applications and will be under ongoing development.
A method of determining popular opinion about something such a product or a political candidate, sentiment analysis attempts to extrapolate whether a text gives a positive, negative, or neutral narrative of an entity.
Joe is an individual programmer and sole owner of Jemigo, his NLP software. For the past 14 years, he's been learning either computer or human languages in some form. Now he's bringing them together, despite having no formal education in Computational Linguistics. He researches and meditates on the topic daily to find new and creative ways to equip Jemigo with natural language capabilities. He challenges everything, obsesses over details, and has the vision to overcome the challenges of NLP.
Intelligence can't be created and the human brain is not fully understood. There is no wisdom in trying to model artificial intelligence on something that we don't comprehend. NLP, is the most important subfield of artificial intelligence. It should be approached pragmatically using modern technologies.
Logic. The key to intelligent software is understanding our own reasoning. A system based on probability is not granular enough to be a good model for natural language. Jemigo is rule-based and uses a custom ontology and dictionary.
Complete Jemigo within the next 2 - 3 years and assist others in harnessing its power in every field to enrich human life.
Enrich human life by endowing software with natural language.
Joe has been a missionary, a soldier, and an Arabic Linguist. Now he is a husband, father, and Software Engineer.
Someone with lots of grit, vision, and an entrepreneurial spirit.
Someone who can write code that will change the world.
"The limits of my language are the limits of my world."
"Knowledge of languages is the doorway to wisdom."
"AI doesn't need to destroy everything in its path, but enhance the human experience as it is."