Searching electronic documents based on example-based search query

A system for searching electronic documents, comprising:

• one or more processors arranged to:
• receive an input representing an example-based search query, the input including one or more sentences or paragraphs of text;
• process the input, including
• identifying one or more terms from the input based on comparing the input with a text database to identify one or more terms in the input contained in the text database;
• determining a weighting factor for each of the one or more identified terms, wherein the weighting factor is a term frequency-inverse document frequency factor determined based on the input and a plurality of electronic documents;
• determining a textual vector of the input based on the one or more determined weighting factors;
• and
• determining, using a topic modeling method based on a Latent Dirichlet Allocation model, a topical distribution associated with the input, including quantizing the topical distribution associated with the input as a topical vector;
• determine, for each of the plurality of electronic documents, a textual similarity score between the input and a corresponding electronic document based on the determined textual vector of the input and a textual vector of the corresponding electronic document;
• determine, for each of the plurality of electronic documents, a topical similarity score between the input and the corresponding electronic document based on the topical vector of the input and a topical vector of the corresponding electronic document;
• determine, for each of the plurality of electronic documents, a topical based linkage score between the input and the corresponding electronic document based on the determined topical similarity scores and linkage relationship in a linkage network of the plurality of electronic documents;
• determine, for each of a plurality of electronic documents, a relevance score between the input and the corresponding electronic document based on
• (i) the determined textual similarity score between the input and the corresponding electronic document,
• (ii) the determined topical similarity score between the input and the corresponding electronic document, and
• (iii) the determined topical based linkage score between the input and the corresponding electronic document;
• determine, based on all each of the determined relevance scores of the plurality of electronic documents, a search result to be provided to the user, the search result containing one or more of the plurality of electronic documents;
• and
• display the search result to the user.

A non-transitory computer readable medium storing computer instructions that when executed by one or more processors, the causes the one or more processors to perform a method for searching electronic documents, the method comprising:

• receiving, from a user, an input representing an example-based search query, the input including one or more sentences or paragraphs of text;
• processing the input, including
• identifying one or more terms from the input based on comparing the input with a text database to identify one or more terms in the input contained in the text database;
• determining a weighting factor for each of the one or more identified terms, wherein the weighting factor is a term frequency-inverse document frequency factor determined based on the input and a plurality of electronic documents;
• determining a textual vector of the input based on the one or more determined weighting factors;
• and
• determining, using a topic modelling method based on a Latent Dirichlet Allocation model, a topical distribution associated with the input, including quantizing the topical distribution associated with the input as a topical vector;
• determining, for each of the plurality of electronic documents, a textual similarity score between the input and a corresponding electronic document based on the determined textual vector of the input and a textual vector of the corresponding electronic document;
• determining, for each of the plurality of electronic documents, a topical similarity score between the input and the corresponding electronic document based on the topical vector of the input and a topical vector of the corresponding electronic document;
• determining, for each of the plurality of electronic documents, a topical based linkage score between the input and the corresponding electronic document based on the determined topical similarity scores and linkage relationship in a linkage network of the plurality of electronic documents;
• determining, for each of a plurality of electronic documents, a relevance score between the input and the corresponding electronic document based on
• (i) the determined textual similarity score between the input and the corresponding electronic document,
• (ii) the determined topical similarity score between the input and the corresponding electronic document, and
• (iii) the determined topical based linkage score between the input and the corresponding electronic document;
• determining, based on each of the determined relevance scores of the plurality of electronic documents, a search result to be provided to the user, the search result containing one or more of the plurality of electronic documents;
• and
• providing the search result to the user, including displaying the search result to the user.