Make sense of

Understanding sentence-level is key when you want to query large quantities of policy documents, contracts, or annual reports. By understanding these documents on a sentence-level you can search through millions of documents in a matter of seconds just by asking a question.

The key challenge

Find coherence between the user search query and the location in the document where an exact answer is given on this question.

Q&A document search

With sequence models you can search in documents by asking a question in natural language. This method goes beyond keyword search and looks much deeper to the meaning of sentences. Similar to how word embeddings are used to find coherent words, for example, a woman is to a man just like a queen is to a king. With sequence models, coherence between an input sentence (question) and an output sentence (answer) is found.

Entity content extraction

Named entity recognition (NER) is used to extract descriptive information from identified entities such as person names, locations, organizations, and numerical and currency expressions. This is particularly relevant if you want to improve your document search and make a distinction between references to ordinary words versus those used as part of a name.

Aim to apply latest machine learning models in real-life?