Surety is performing system maintenance this weekend. Electronic date stamps on new Prior Art Database disclosures may be delayed.
Browse Prior Art Database

Tone amendement system for text input of non native speakers

IP.com Disclosure Number: IPCOM000247848D
Publication Date: 2016-Oct-06
Document File: 3 page(s) / 40K

Publishing Venue

The IP.com Prior Art Database


Sentence tone amendement system for text based communications across native and non native speakers

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

Page 01 of 3

Tone amendement system for text input of non native speakers

Software systems for analysis of the 'tone' of a spoken or written sentence are becoming both commonly developed and popular.This article is about a system to prevent and reduce the risk of mis-classification of the tone of a sentence particularly for non-native speakers. Natural language classifiers for Tone Analysis on the market are often based on statistical training of large corpora of texts from different sources. However non native speakers have usually a limited knowledge of the a certain language (eg. English if they are speakers from EU continental countries) with respect to a native speaker, this applies to the cardinality of the word dictionary or to idiomatic expressions or slang expressions. This can lead to a imperfect communications between two parties involved in a text based communication over a network (chat, email, social platforms), this can happen even if the parties are both non native or if one of the party is native and the other party is not native, even if the kind of problems could be different in the two cases (a native speakers can notice a bad tone in cases where a non-native speaker can ignore that fact).

Followin is the overall architecure of a system that prevents unwanted wrong tone to be submitted in a text communication system. Using state of the art language technologies tools to identify whether the text being written is native or not.

In the latter case a Tone Analysis is done on the input text and the outcome is presented to the user with possible suggestion from an internal knowledge base of similar cases, if the user simply confirm the text he just did input nothing happen and the orignal text is submitted. Differently the system highlights the affected sentences tagged with the identified tones and the proposed correction and related corrected tones.

If the user is satisfied with the proposed correction he simply accepts the co...