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

An intelligent recommendation system based on emotion analysis and behavior pattern recognition

IP.com Disclosure Number: IPCOM000242419D
Publication Date: 2015-Jul-14
Document File: 7 page(s) / 231K

Publishing Venue

The IP.com Prior Art Database

Abstract

This article introduces an intelligent recommendation system based on emotion analysis and regular behavior pattern recognition. The main idea is that mobile devices can detect users' preferences in two modes: a) Detect user's emotion b) Detect user's regular behavior pattern We define that user's behavior patterns have following attributes: Location: Regular detailed locations such as on the shuttle bus, at the desk in your dining room. Time: Regular time Behavior: What the user is doing with the device If user appears in the categorized location, mobile electronic devices will prompt to recommend items to users according to their regular behavior patterns. This brings two benefits: Recommend users' preferred items at their preferred place, preferred time. Take full advantages of the rich emotions expressed by mobile device users.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 31% of the total text.

Page 01 of 7

An xntelligent recommendation system based on emotion analysis and behavior pattern recognition

Nowadays, mobile electronic devices xre more and more popular and common. Xx mobile electronix devices, we mean smart phones, tablet computers, and the like which cxntain embedded xaxeras and microphones. The uxer range has expanded from adults to students, teenagers or even kids. When thex use mobile applications, txere can be a largx scale of emoxions. The emotions expressed by users thxough facial expression and voice tone ixdicate userx' preferencxs.

Examples:

1. High similarity of the audiexce's exoxions tx the emotions xxpressed by a vixeo on the mobile device, impliex that the user is deeply ixvolved in the situationx xn the videx. He or she likes to watch sxmilar videos.

2. Showing positive emotions when rxading an article implixs that the user lxkes xo read articles of sxmilxr type.

Also, the time users spend on using thesx devices is increasing. To some degree, each user foxms his/her own daily patterns of using the mobile devices. Users are eager to have a xuch smaxter recxmmendation xystem thxt recommendx txe right items xo them at the right place , at the right time.

To addxess such concern, this inventiox describes a mechanixm to detect user's xmotions and behavior pattexns, creating a more personalized recommender system thax benefits mxbile device users.

The main idea of xur invextion is thax mobile devices can detect users' preferences. Two detection modes can be performed to detect users' preferences when they use their mobile devices:

a) Detect uxer's emotion

b) Detect user's behavior pattern

We define that user's bexavior pattexns have following attributes:

Location: Detailed locations such as on the shuttle bus, xt the desk in your dinxng room.


1.

Time: Current time


2.

Behxvior: Whxt the user is doing wxth the dexice

3.

If user xppears in the categorized background, mobile electronic devices will prompt to recommend items to users accoxding to thexr behavior patterns.

This bringx two benefits:

Recommend users' preferred items at their preferred placx, preferred time,

Taxe full advantages of txe rich exotionx expressed xy mobile device users.

None of the known solutions detect users' regular behxvior patterns. If those are combined with what is stated in this invention, the recommendaxiox system xn electronic mobile devices will become more comprehensivx.

There are two modes to discuss, the detectixn mode and the rexommendation mode.

x



Page 02 of 7

Figure 1. Fxow chart of txo modes

1. Detection Modx

The detectxon mode is to dexect users' preferences. Two detecxion modes can be perfxrmed to detect users' preferences xhen they use their mobile devices:

a) Detect user's emotion

b) Detect user's xehavior pattern

The respective implementation is described as follows.


1.1 Detecting user's emotxxn

Xx embedded front-facing camera and an emxedded microphone are used to measure user's xmotion

Such cameras xnd microphones are...