Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Multi-lingual hash tags extraction and suggestion for social media messages

IP.com Disclosure Number: IPCOM000246534D
Publication Date: 2016-Jun-16
Document File: 4 page(s) / 96K

Publishing Venue

The IP.com Prior Art Database

Abstract

While reading the posts via social networking, the people could utilize hash tags to find the exact same keyword with the same language of other posts. However, sometime people also would like to find the keyword with different languages of other posts to read more variable information. The proposed idea is to make the extracted keyword (hash tags) related to other translations by machine translation. So that user can display/aggregate the document with translated keywords. Also, the mehod proposes a method to save the storage size of saving keywords connection for all languages by connecting all languages to the baseline language.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 4

Multi-lingual hash tags extraction and suggestion for social media messages

While reading the posts via social networking, the people could utilize hash tags to find the exact same keyword with the same language of other posts. However, sometime people also would like to find the keyword with different languages of other posts to read more variable information. For example, the people who love the bakery to utilize the "cupcake" hash tag in Chinese to find the information, but actually they also would like to search the information with "cupcake" hash tags in English. Moreover, if the author of posts did not make the keyword for the posts, these useful information might be ignored when people use the hash tag to find the related information.

Core idea and claims:
1. By connecting all languages to the baseline language, we can save the storage size of saving keywords connection for all languages. ( Complexity N instead of N*N)

2. For display/aggregation result improvement, the proposed method makes the extracted keyword (hash tags) related to other translations by machine translation. So that user can display/aggregate the document with translated keywords.

3. For hash tag suggestion, when using tag functions on the social media or other supported system, it will provide client with the suggested tag by the extracted keywords(hash tags) and associate to item 1)

Advantage:
1. User can get all desired language display/aggregate results by inputting a keyword (hash tag)

2. User will get suggested hash tag names after it finished a post
3. For server side, it only needs do store the extracted keywords (hash tags) and the translation connection after keyword (hash tags) extraction.
4. Save storage size by connecting all languages to the baseline language

Workflow for posting an article:


1. A user posts an article in language A to the website.

2. The website extracts the keywords of the article by the keyword extractor. ( Keywords could be extracted by text mining which is a background art here)


3. The website translates the keywords into a baseline language.

4. The website saves the article, keywords in language A and keywords in baseline language into the data store.

Every keyword (hash tag) is translated and will be compared in the baseline language. This simplifies the keywords (hash tags) display/aggregation across multiple languages.

1



Page 02 of 4

Workflow for search/display articles by keywords (hash tags):


1. A reader inputs the query keywords (hash tags) in language B in the website to display/aggregrate interested articles.

2. The website translates the keywords (hash tags) into the baseline language.

3. The website queries the data store with the keywords (hash tags) in both language B for articles in language B and keywords (hash tags) in the baseline language for articles written in other languages. Also, the su...