Machine Learning is being increasingly used in several business applications. There are many tools that can be used to solve a wide variety of ML problems. However, in many cases the problem that one is solving is specific to the business and one can’t use the trained models that are available off the shelf. One of the major challenges in any machine learning problem is to train the model using data. It is an iterative process and requires insight into the workings of the algorithms as well as the data to get the model right which in turn is essential to make accurate predictions to solve specific problems. This applies for NLP tools as well.
We have worked with several popular ML tools and libraries including Weka, Apache Mahout, Spark MLLib and Scikit-Learn. We have also worked with various NLP libraries including OpenNLP, StanNLP and nltk. We have a deep understanding of how to use these tools and build applications from scratch. We have expertise in a number of NLP techniques like Information retrieval (Lucene/Solr), Noun Phrase Analysis, Concept/Knowledge extraction, Named Entity Recognition. The fact that our team members have strong Maths background from top tier engineering institutes also helps. In addition to this, we also have expertise in data annotation to train the models. We have a team of people who are strong in English to annotate text data for NLP problems.
Experience working with popular ML tools & libraries like Weka, Mahout, MLLib, Scikit-Learn
Experience working with NLP libraries like OpenNLP, StanNLP and nltk
Expertise in NLP techniques like information retrieval, Noun Phrase Analysis, Concept/Knowledge extraction, Named Entity Recognition
Strong background in Mathematics and ability to train models
Experience in creating training data, annotating text data for NLP problems