Method and system for automatic summarization and digest of celebrity news
A system and method for automatically creating a digest of celebrity information received over an adjustable time range from publicly available electronic data sources by applying automated qualitative and quantitative analytical methods to select information for inclusion in the digest and to convert such information to summary form.
1. A method of summarizing information concerning celebrities, comprising a plurality of steps of:
establishing a relational database for holding such information
wherein such information contains data selected from the group consisting of:
name;
gender; and
age;
gathering one or more stories from a plurality of sources;
parsing the stories to determine specific indicators of one or more celebrities;
storing said stories in said relational database with appropriate tags to enable retrieval by a user of the database;
automatically summarizing the stories based on story text for each celebrity; and
presenting said story summaries for viewing by said user;
wherein the step of summarizing said stories further comprises a plurality of steps of:
generating a corpus of data formed by concatenating all story text associated with a selected celebrity;
stripping predetermined incidental words from said corpus of data;
generating a plurality of bigrams of the remaining text in the corpus of data;
performing term frequency-inverse document frequency analysis on the plurality of bigrams;
assigning a weight value to each bigram; and
selecting one or more sentences from said corpus of data based on the weighted value of said bigrams.
2. The method of claim 1 , wherein said stories are gathered over a global communication network.
3. The method of claim 1 , wherein the step of parsing the stories further comprises the steps of:
tagging each story based on date, title, and abstract information;
matching patterns in the stories to a predetermined list of known celebrities;
identifying celebrity names using domain specific terminology;
identifying keywords in the story indicative of celebrity information.
4. The method of claim 1 , further comprising a plurality of steps of:
generating a corpus of data formed by concatenating all story text associated with a selected celebrity;
parsing said corpus of data into parts of speech that can be assigned qualitative values;
assigning a qualitative value to each said part of speech;
performing term frequency analysis on said corpus of data; and
assigning a weighted value to said data based on said qualitative values for each part of speech.