Team discovery for community-based question answering systems

An information handling system comprising:

• one or more processors;
• a memory coupled to at least one of the processors;
• a set of computer program instructions stored in the memory and executed by at least one of the processors in order to perform actions of:
• defining a set of teams in a Community-Based Question/Answer (CQA) network based on at least one question from a set of questions and at least one expert from a set of experts;
• determining a set of offset vectors of each of the set of teams based on an amount of the set of experts in their corresponding team;
• computing a set of embedding vectors comprising a set of team center vectors, a set of expert embedding vectors, and a set of question embedding vectors based on the set of teams, the set of experts, and the set of questions;
• computing a set of loss functions based on the set of offset vectors, the set of team center vectors, the set of expert embedding vectors, and the set of question embedding vectors;
• mapping a set of question nodes and a set of expert nodes to an embedding space based on the set of embedding vectors and the set of loss functions, wherein the set of question nodes corresponds to the set of questions and the set of expert nodes corresponds to the set of experts in the CQA network;
• mapping a new question to the embedding space based on analyzing the new question against the set of questions;
• selecting a subset of the set of expert nodes based on their location in the embedding space relative to the mapping of the new question;
• and
• transmitting the new question to a subset of the set of experts that correspond to the subset of expert nodes.

A computer program product stored in a computer readable storage medium, comprising computer program code that, when executed by an information handling system, causes the information handling system to perform actions comprising:

• defining a set of teams in a Community-Based Question/Answer (CQA) network based on at least one question from a set of questions and at least one expert from a set of experts;
• determining a set of offset vectors of each of the set of teams based on an amount of the set of experts in their corresponding team;
• computing a set of embedding vectors comprising a set of team center vectors, a set of expert embedding vectors, and a set of question embedding vectors based on the set of teams, the set of experts, and the set of questions;
• computing a set of loss functions based on the set of offset vectors, the set of team center vectors, the set of expert embedding vectors, and the set of question embedding vectors;
• mapping a set of question nodes and a set of expert nodes to an embedding space based on the set of embedding vectors and the set of loss functions, wherein the set of question nodes corresponds to the set of questions and the set of expert nodes corresponds to the set of experts in the CQA network;
• mapping a new question to the embedding space based on analyzing the new question against the set of questions;
• selecting a subset of the set of expert nodes based on their location in the embedding space relative to the mapping of the new question;
• and
• transmitting the new question to a subset of the set of experts that correspond to the subset of expert nodes.