Query expansion using a graph of question and answer vocabulary

A computer-implemented method for query expansion using a graph of question and answer vocabulary, the method comprising:

• providing a set of training data in a given domain in the form of training question texts and training answer texts;
• identifying disjoint answer words in a training answer text that do not occur in an associated training question text;
• adding nodes to a graph as question word nodes and answer word nodes, wherein the answer word nodes are for the identified disjoint answer words in the set of training data;
• adding edges to the graph between a disjoint pair of a question word in a training question and a disjoint answer word in an associated training answer;
• and
• providing weightings for the nodes and edges based on frequency data of words in the set of training data.

A system for query expansion using a graph of question and answer vocabulary, the system comprising:

• a processor;
• and
• a memory communicatively coupled to the processor and configured to store computer program instructions;
• wherein the processor is configured to execute the computer program instructions to:
• provide a graph of question word nodes and answer word nodes generated from a set of training data for a given domain in the form of training question texts and training answer texts, wherein the answer word nodes are of disjoint words that do not occur in both a training question text and an associated training answer text, and edges are provided between a disjoint pair of a question word node for a question word in a training question and an answer word node for a disjoint answer word in an associated training answer;
• provide weightings for the nodes and edges based on frequency data of words in the set of training data;
• receive a user query input;
• activate input nodes in the graph for at least some words in the user query input;
• apply spreading activation through the graph using the weightings;
• and
• return a result of a top n most highly activated nodes that are output as candidate words for expansion of the user query input.

A system for query expansion using a graph of question and answer vocabulary, the system comprising:

• a processor;
• and
• a memory communicatively coupled to the processor and configured to store computer program instructions;
• wherein the processor is configured to execute the computer program instructions to:
• provide a set of training data in a given domain in the form of training question texts and training answer texts;
• identify disjoint answer words in a training answer text that do not occur in both an associated training question text;
• add nodes to a graph as question word nodes and answer word nodes, wherein the answer word nodes are for the identified disjoint answer words in the set of training data;
• add edges between a disjoint pair of a question word in a training question and a disjoint answer word in an associated training answer;
• and
• provide weightings for the nodes and edges based on frequency data of the words in the set of training data.

A computer program product for query expansion using a graph of question and answer vocabulary, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:

• provide a graph of question word nodes and answer word nodes generated from a set of training data for a given domain in the form of training question texts and training answer texts, wherein the answer word nodes are of disjoint words that do not occur in both a training question text and an associated training answer text, and edges are provided between a disjoint pair of a question word node for a question word in a training question and an answer word node for a disjoint answer word in an associated training answer;
• provide weightings for the nodes and edges based on frequency data of the words in the set of training data;
• receive a user query input;
• activate input nodes in the graph for at least some words in the user query input;
• and
• apply spreading activation through the graph using the weightings to result in a top n most highly activated nodes that are used as candidate words for expansion of the user query input.