Both question types, semi-open as well as fully-open, are a subset of open-ended questions (as opposed to closed questions like multiple-choice).
Fully-open questions leave room for a broad spectrum of answers, often including sentiment & emotions and usually have people respond in sentences or phrases.
Semi-open / list questions are narrower in their scope, closer to keywords and often deliver a list or an enumeration.
Caplena uses two different solutions to analyze these two types.
Why are there two different solutions to process these questions?
Both question types have a different set of challenges. Having two separate solutions ensures the most efficient and accurate analysis. The main challenge while analyzing semi-open questions lies in spelling mistakes and different ways of separation. Fully-open questions on the other hand, are more difficult to analyze and have a very broad set of challenges. Some of them being the usage of different expressions or concepts for similar topics, as well as the interpretation of sentiment and emotions.
Example question: What did you like or dislike about your recent trip to Paris?
Example response: Enjoyed the fantastic food in those sweet bistros, although the waiters were at times a bit grumpy.
Response format: Sentence fragments or even multiple sentences including sentiment / emotions.
Purpose: Broad set of use-cases, e.g. NPS, product reviews, social media posts.
Caplena analysis: The AI analyzes meaning and content-related concepts, including negations, synonyms and emotions.
Example question: What car brands are you familiar with?
Example response: Mercedes, Ford and Toyota
Response format: Mostly single words or short word groups, many duplicates.
Purpose: Mostly used in an unaided awareness and/or top-of-mind context.
Caplena analysis: The AI considers the similarity of spellings.