Natural Language Processing

The adoption of NLP in our framework allows us to support flexibility in dialogues. Thus, we selected the NLP library OpenCCG, which is based on CCG grammar formalism. Currently, we are making progress in the development of an Italian grammar for the hypertensive patients management by using OpenCCG. This is an example of the derivation of the sentence “ho una pressione massima di 110” (“my systolic blood pressure is 110”) together with the semantic recognition.

In this example we can see the syntactic categories and the semantic representation. The categories of each part of the speech are shown in the first row below the sentence. For instance, the determinant “una”(“a”) is formed by a noun to the right of the slash and the result category is a noun phrase. Since “pressione” is a noun and is to the right of the determinant “una”, they are derived by the forward composition rule >, as shown in the second row down the sentence. On the other hand, in the semantical representation a1 is a discourse referent for the event of having “avere”, which takes place in the present. It is related to x1, the discourse referent for the patient, by the Owner role, and to p1, the discourse referent for the blood pressure, via the Condition role. The referent p1 is in turn related to m1, the discourse referent for the systolic adjective “massima”. Also, p1 is related to n1 the blood pressure measure by the measure role.