NLP edge case - hyphen separated words aren't highlighted as one word (fifty- two)

[jsms90]

For some reason there is often a space between the hyphen and the second word. Sometimes a space on both sides (?)

Sometimes the first word is picked up by compromise (the NLP library), sometimes just the second. Sometimes both, but separately from each other.

With a two digit word like fifty-two, users could click on both because clicking 50 & clicking will give a combined total of 52 👍 But this doesn't work well for 3 digit numbers.

[jsms90]

First attempt:

if (this.abstract) {
        const abstractWithWhitespaces = this.abstract
          .replace(/([a-zA-Z]+[\,\.\)\}\]\!\?]+)(\d)/g, "$1 $2")

          .replace(/(\d+[\,\.\)\}\]\!\?]+)([a-zA-Z]+)/g, "$1 $2")
          .replace(/([a-zA-Z]+)\s?\=\s?(\d)/g, "$1 = $2");
        let nlpExtractedNumbers = nlp(abstractWithWhitespaces).match("#Value").out("text");
        let parsedAbstract = abstractWithWhitespaces
        while (nlpExtractedNumbers.match(/([a-zA-Z]+)\s?\-\s([a-zA-Z]+)/)) {
          const wordsWithDashes = /([a-zA-Z]+)\s?\-\s([a-zA-Z]+)/.exec(nlpExtractedNumbers);
          parsedAbstract = abstractWithWhitespaces.replace( wordsWithDashes[0], wordsWithDashes[1] + wordsWithDashes[2]);
          nlpExtractedNumbers = nlpExtractedNumbers.replace( wordsWithDashes[0], wordsWithDashes[1] + wordsWithDashes[2]);
        }
        return nlp(parsedAbstract).out("html");
      }

But we still have them showing up as separate words

Narrowing it down, parsedAbstract gives ninety-five as one word:

' Event- related  spectral  perturbations  (ERSPs;  event- related  mean  power  spectral
changes)  and  inter- trial  coherence  (ITCs;  event- related  consistency  of  spectral  phase)
reveal  a  more  comprehensive  overview  of  EEG  activity.  Ninety-five  subjects  (56  MS
patients,  39  controls)  completed  visual  and  auditory  two- stimulus  P3b  event- related
potential  tasks  and  the  PASAT. '

But nlp(parsedAbstract).out("html") treats ninety and five as 2 separate words and puts them in separate spans with different classes, hence the separate styling:

<span
    class="nl-TitleCase nl-Hyphenated nl-Cardinal nl-Value nl-TextValue"
>Ninety</span>-
<span
    class="nl-Hyphenated nl-Cardinal nl-Value nl-TextValue"
>five</span>