Interactive intent modeling
Interactive Intent Modeling describes search methods that allow users to explore information in an explorative way. It is used in information retrieval systems and promotes research and finding of information that the user needs. Such a system was developed, for example, by the researchers at SciNet , which contains a collection of 50 million scientific articles that can be explored and accessed using an intent modeling system.
background
Search engines nowadays have the problem that they do not support an exploratory search process. Users find it difficult to search for their desired resource. This is where the vocabulary mismatch problem occurs. This means that people use different terms and expressions for one and the same subject. Because of this, people waste a lot of time and energy trying to find the right words for their search query instead of working with the actual resource. Another quality of people is that they are better at recognizing information than at remembering it. This phenomenon also occurs as déjà-vu , which makes you feel like you've seen something before.
functionality
An Interactive Intent Modeling System processes the user's search query and uses it to calculate new terms, which it then presents to the user as suggestions. These suggestions, also called intents, are distributed on a radar according to their relevance, with the inner radius containing intents that are similar to the search term and the outer circle containing intents that are little similar to the original search term, but in some way through the same Categories or subject areas are related. The user now has the option of interacting with these intentions: either the user selects an intent as a new search term or sorts the intentions according to their relevance. The closer an intent was moved to the center of the radar, the more relevant it is for the user.
The system generates a new result from this new arrangement, with which the user can also continue to interact. This creates an iterative-exploratory process of information acquisition.
Individual evidence
- ↑ a b Interactive intent modeling: information discovery beyond search Article from the ACM Digital Library. Retrieved May 10, 2017
- ↑ Directing exploratory search: reinforcement learning from user interactions with keywords Article from the ACM Digital Library. Retrieved May 10, 2017
- ^ The vocabulary problem in human-system communication Article from the ACM Digital Library. Retrieved May 10, 2017
- ↑ Learning and memory: An integrated approach ( Memento of the original from June 23, 2017 in the Internet Archive ) Info: The archive link was automatically inserted and not yet checked. Please check the original and archive link according to the instructions and then remove this notice. Article from the TU Darmstadt University Library. Retrieved May 10, 2017
