Browse Prior Art Database

A Method and System for Highlighting Topics in Online Conversations

IP.com Disclosure Number: IPCOM000240222D
Publication Date: 2015-Jan-14
Document File: 2 page(s) / 35K

Publishing Venue

The IP.com Prior Art Database

Related People

Vidit Jain: INVENTOR [+4]

Abstract

A method and system is disclosed for highlighting topics in online conversations. The method and system automatically discovers the different facets in a message thread and highlights the important content extracted from one of the messages. A collection of the highlighted content is referred to as topic highlights. The topic highlights enables the users to quickly catch up with the different facets of conversations in a topic.

This text was extracted from a Microsoft Word document.
This is the abbreviated version, containing approximately 52% of the total text.

A Method and System for Highlighting Topics in Online Conversations

Abstract

A method and system is disclosed for highlighting topics in online conversations.  The method and system automatically discovers the different facets in a message thread and highlights the important content extracted from one of the messages.  A collection of the highlighted content is referred to as topic highlights.  The topic highlights enables the users to quickly catch up with the different facets of conversations in a topic.

Description

Online communities allow users to join a group and participate in conversations to collaborate on a several formal and informal topics over a period of time.  Most online communities consider each message thread in the conversations as a separate topic.  The message thread can be an email thread or a conversation thread.  Most group members participate in the conversations on various topics while some of the group members only follow the conversations.  Popular topics typically attract a large number of posts.  There can be difficulty for members to follow the popular topics in the conversations which include multiple divergent facets.  Generally, group members may be interested in only a subset of the facets in the conversations.

Disclosed is a method and system for highlighting topics in online conversations. 

The method and system automatically discovers the different facets in a message thread and highlights the important content extracted from one of the messages.  A collection of the highlighted content is referred to as topic highlights.  The topic highlights enables the users to quickly catch up with the different facets of conversations in a topic.  Further, appropriate instrumentation of overall audience for different facets can provide additional signals for personalization and monetization of online communities.

In one implementation of the method and system, each message belonging to the topic is analyzed to identify the original contribution.  There are several pre-defined rules that are applied to the message to remove the salutations, header and footer, replied-to, and forwarded content.  The remaining content in the message is referred to as the "message document", hereafter.  The method and system sets the number of topic highlights as a constant K (e.g., typical value of 3 to 5) based on the number of distinct messages belonging to the topic.  A latent Dirichlet allocation model learns all message documents for the given topic as the learning corpus, and K as the parameter that specify the desired number of statistical “topics”.  The K statistical "topics" correspond to K facets of the conversation, and is represented as a probability distribution over the entire vocabulary.  For each facet, each word in the vocabular...