System and method for automatic generation of information-rich content from multiple microblogs, each microblog containing only sparse information

• US 9,990,368 B2
• Filed: 08/06/2012
• Issued: 06/05/2018
• Est. Priority Date: 07/04/2012
• Status: Active Grant

• Current Assignee: Qatar Foundation
• Original Assignee: Qatar Foundation
• Inventors: Walid, Magdy, Darwish, Kareem, Ali, Ahmed
• Primary Examiner(s): To, Baoquoc N
• Application Number:
  US14/408,732
• Publication Number:
  US 20150261773A1
• Time in Patent Office: 2,129 Days
• Field of Search: 707E17095, 70799901, 707705
• CPC Class Codes: G06F 16/148 File search processing
• CPC Class Codes: G06F 16/345 Summarisation for human users
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06Q 50/01 Social networking

Abstract

A system and method for automatic generation of information-rich content from multiple microblogs, each microblog containing only sparse information, the method comprising collecting a population of microblogs comprising microblog data, each microblog containing a limited number of characters; providing a user interface allowing entry of a search query; matching a search query entered on the user interface to data in the microblog data; providing the results of the matching process as a sub-set of microblog data; applying processing techniques to the sub-set of microblog data; and generating a summary report of the processed sub-set of microblog data.

Claims

1. A method for automatic generation of information-rich content from multiple microblogs, each microblog containing only sparse information, the method comprising:

collecting a population of microblogs comprising microblog data, each microblog containing a limited number of characters;

providing a user interface allowing entry of a search query;

matching a search query entered on the user interface to data in the microblog data, the microblog data based partially on contributor mood;

providing the results of the matching process as a sub-set of microblog data;

applying processing techniques to the sub-set of microblog data;

wherein processing techniques include;

using natural language processing techniques to generate at least one category in the sub-set of microblog data, the category based on a link between a sentiment and the sparse information of the sub-set of microblog data;

monitoring microblog contributor mood for the matched search query and observe change over time;

generating contributor mood data for an event in the sub-set of microblog data over time based on a microblog and the category, wherein the contributor mood is associated with the category based on a sentiment in the sparse information of the microblog; and

outputting a group of microblogs based on the sub-set of microblog data associated with the contributor mood data of the event, wherein each microblog is based on a correlation between the matched search query and the sentiment associated with the contributor mood data.

2. The method according to claim 1, wherein the sentiment is one of a set of sentiments including at least one of positive, negative, funny, sad, happy, associated with a mood category in the microblogs, the method comprising:

dividing the sub-set of microblog data into different categories of microblog based on the sentiment; and

incorporating results for each of the different categories in the summary report.

3. The method according to claim 2, wherein the category of microblog is selected from at least one of the following:

microblogs containing the matched search query;

microblogs containing sentiment;

microblogs containing positive/negative sentiments, wherein the positive/negative sentiments include a subset of sentiments in each of the positive/negative sentiments;

microblogs containing other forms of sentiment;

microblogs containing sarcastic posts and/or joke-like posts related to the search query;

microblogs containing links related to the search query; and

microblogs containing video links related to the search query.

4. The method according to claim 3, wherein the natural language processing is selected from at least one of the following:

text normalization

named entity recognition

keyword/key-phrase extraction

sentiment analysis.

5. The method according claim 4, wherein the processing comprises identifying URLs and expanding URLs.

6. The method according to claim 4, wherein the summary report provides at least one of the following:

a relationship plot between information need and the extracted items;

navigation of the results over a user-selected time window;

extracted top videos and links related to such events/entity;

extracted top sarcastic posts and jokes-like tweets for the required event; and

extracted top positive/negative and other type of categorized sentiments posts.

7. The method according to claim 4, wherein the microblog is selected from the group comprising:

a post on a social networking site;

a comment on a news article;

a comment or a post on a forum;

a comment on a social networking site; and

other user generated content (UGC).

8. The method according to claim 7, wherein the microblogs are normalised into microblog data having a predetermined format.

9. The method according to claim 8, wherein normalised microblog data is stored in an index of microblogs to comprise the population of microblogs.

10. The method according to claim 8, wherein the predetermined format includes at least a selection of the following:

microblog text;

microblog identifier; and

microblog time-stamp.

11. A system for automatic generation of information-rich content from multiple microblogs, each microblog containing only sparse information, the system including:

a computing device having a processor and a memory; and

a storage device;

the computing device being configured to perform the method according to any preceding claim.

12. The system according to claim 11, further including a visual display for displaying an interface to a user, and to receive a search query from a user, such that the input of a search query by the user causes the computing device to output to the interface a summary report of the processed sub-set of microblog data corresponding to the search query.

13. A computer-readable medium storing instructions which when executed to run on a processor cause the processor to perform the steps according to the method of claim 1.

15. The method according to claim 1, wherein the contributor mood is based at least partially on emoticons, wherein the sentiment is one of a set of sentiments including at least one of positive, negative, funny, sad, associated with a mood category in the microblogs.

14. A system for automatically generating information-rich content from multiple microblogs, each microblog containing only sparse information, the system comprising:

a collection of microblogs comprising microblog data, each microblog containing a limited number of characters, the microblog data based at least partially on contributor mood;

a user interface allowing entry of a search query; and

a hardware processor configured to;

match a search query entered on the user interface to data in the microblog data;

generate at least one category in the matched microblog data, the category based on a link between a sentiment and the sparse information of the matched microblog data;

monitor microblog contributor mood for the matched search query and observe change over time generating contributor mood data for an event in the sub-set of microblog data over time based on a microblog and the category, wherein the contributor mood is associated with the category based on a sentiment in the sparse information of the microblog; and

output a group of microblogs based on the sub-set of microblog data associated with the contributor mood data of the event, wherein each microblog is based on a correlation between the matched search query and the sentiment associated with the contributor mood data.