| CPC G06F 16/93 (2019.01) [G06F 3/0482 (2013.01); G06F 16/951 (2019.01); G06V 30/40 (2022.01); G06F 2218/12 (2023.01)] | 13 Claims |
|
1. A machine-implemented method for world-wide web document searching, identification, classification, event scoring, generating a graphical user interface (GUI) for a research system, and implementing a detection system for identifying an increase in a number of documents pertaining to a particular category of multiple categories over a particular period of time, the method comprising:
one or more processors; and
a memory, coupled to the one or more processors, that stores code that when executed by the one or more processors causes the one or processors to perform operations comprising:
executing world wide web searches with a webcrawler on a regular, automated basis to (i) detect new documents and automatically retrieve the documents, (ii) to enable detection of when events have occurred for categories in the detection system, and (iii) to enable classification of content in the documents as relevant to the categories wherein the documents include content elements and structural elements;
automatically analyzing the documents using category models that score the documents for relevancy to a particular category to classify the documents, wherein each model represents a specific entity and includes data that is used to identify documents related to the entity that the model represents, each model includes word association patterns of document elements associated with the score of the documents and includes parameters used in the analysis of documents by the model, wherein automatically analyzing the document using the category models comprises:
applying the models of each of the documents by identifying the word association patterns of document elements in the document;
for each document, calculating a relevance score for each document's relation to each of the categories represented by the models by applying the word association patterns identified in the document and associated scores of the documents; and
for each document, classifying the document by assigning the relevance score to each document for one or more of the categories over a particular period of time;
based on the classification of the documents, for each particular category, identifying a first volume of documents relevant to the particular category over the particular period of time;
removing duplicate documents from the first volume of documents;
determining if the first volume of documents exceeds an average number of documents relevant to the category for a background time period;
assigning an event score to the first volume of documents when the first volume of documents exceeds the average number of documents relevant to the category for a background time period;
normalizing the event scores across a plurality of the categories;
comparing the normalized event scores across the plurality of the categories;
determining that an event has occurred when the normalized event score exceeds a particular threshold;
receiving a request from a user of the research system for information about at least one of the particular categories;
generating a chart that displays a set of events associated with the particular category over a particular period of time and displays a normalized comparison of events across the plurality of the categories, wherein the set of events comprises events appearing on the world wide web relevant to the particular category that have an assigned normalized event score that exceeds the particular threshold, and wherein the chart comprises a selectable item for a particular event; and;
incorporating the chart into a GUI for the particular category for transmission to a computer system, wherein the GUI displays the selectable item and upon selection of the selectable item by a user of the computer system causes the GUI to display a set of documents appearing on the world wide web at a range of time corresponding to the particular event.
|