Imagine trying to form a sentence out of these words
- JP Morgan
- hair dressing
- capital markets
- car mechanics
- on-the-job training
- deal making
That actually is one of the problems with Natural Language Processing.
A possible sentence is as follows:
JPMorgan is hiring its first investment banking apprentices in the UK, applying the on-the-job training method previously reserved for hairdressing and car mechanics to roles on its capital markets and dealmaking teams.
One big issue is the following phrase, where we center on the term : capital (next to markets) and look at the surrounding +/- 7 words.
It seems unfortunately that car mechanics are within the window of 7 words as well as the term “to roles”. If a system treated “capital markets and deal making teams” as a single entity then the relation appears that “car mechanics” were hired to JP Morgan’s capital markets and deal making teams.”
These are some of the challenges in performing Natural Language Processing.
More to follow.