Specific embodiment
The invention will be further described with the following Examples.
Referring to Fig. 1, one kind of the present embodiment is capable of the intelligent earphone of autonomous control, including first information acquisition unit 1, the
One information process unit 2, first control unit 3, the second information acquisition unit 4, the second information process unit 5 and the second control are single
Member 6, the first information acquisition unit 1 are used to acquire the wearing information of user, and the first information processing unit 2 is for receiving
The wearing information of the user simultaneously generates the first command signal, and the first control unit 3 is according to the first command signal control
Intelligent earphone processed is switched to operating mode from sleep pattern or is switched to sleep pattern from operating mode, and second information is adopted
Collection unit 4 acquires the identity information of user when being switched to operating mode for intelligent earphone, second information process unit 5 is used
In the identity information for receiving the user and the second command signal is generated, second control unit 6 is according to second instruction
Signal is recommended music to user and is played.
The present embodiment realizes the autonomous control of intelligent earphone, can be according to the wearing situation of user and the identity control of user
The working condition of intelligent earphone processed and the music of broadcasting.
Preferably, second control unit 6 includes scoring acquisition unit, rating matrix unit, neighbour user's selection list
Member and music recommendation unit, the scoring acquisition unit are used for obtaining scoring of the user to music, the rating matrix unit
In determining user to the rating matrix of music according to scoring, neighbour's user selection unit is for searching and user's phase to be recommended
As neighbour user, the music recommendation unit be used for according to neighbour user to user to be recommended recommend music.
The second control unit of this preferred embodiment realizes the personalized recommendation of intelligent earphone music.
Preferably, the rating matrix unit is for determining user to the rating matrix of music:
Rating matrix: EH=[p is determined according to the following formulaik]m×n, EH, p in above-mentioned formulaikIndicate i-th of user uiTo kth
A music xkScoring, EH indicate rating matrix, m indicate user quantity, n indicate music quantity, if user is not to certain
Music scoring is then 0 to deserved scoring, scores higher, illustrate that user is bigger to the fancy grade of music;
This preferred embodiment rating matrix unit establishes user to the rating matrix of music, selects for subsequent neighbour user
Unit searches neighbour user and lays a good foundation.
Preferably, neighbour's user selection unit includes first neighbour's user selection unit, the second neighbour user selection
Unit and comprehensive neighbour's user selection unit, the first neighbour user selection unit is for determining user to be recommended and other use
First similarity at family, the second neighbour user selection unit is for determining that user to be recommended is similar to the second of other users
Degree, comprehensive neighbour's user selection unit are used to determine the neighbour of user to be recommended according to the first similarity and the second similarity
User;
The first neighbour user selection unit is used to determine the first similarity of user to be recommended Yu other users:
If user uiWith user ujThe collection of music to score jointly is Uij, the first similarity is calculated using following formula: In above-mentioned formula, picAnd pjcRespectively indicate user uiWith
User ujTo music xcScoring, piAnd pjRespectively indicate user uiWith user ujTo the average value of the scoring of all music, EM1
(ui,uj) indicate user uiWith user ujThe first similarity;
First similarity value is bigger, then the similitude of user is bigger, conversely, then similitude is smaller.
The second neighbour user selection unit is used to determine the second similarity of user to be recommended Yu other users:
Using each user to the scoring of all music as a n-dimensional vector, if user uiWith user ujScoring vector
It is denoted as I=[p respectivelyi1,pi2,…,pin] and J=[pj1,pj2,…,pjn], all collection of music are X, calculate second using following formula
Similarity:In above-mentioned formula, pikAnd pjkRespectively indicate use
Family uiWith user ujTo music xkScoring, EM2(ui,uj) indicate user uiWith user ujThe second similarity;
Second similarity value is bigger, then the similitude of user is bigger, conversely, then similitude is smaller.
Comprehensive neighbour's user selection unit is used to determine user to be recommended according to the first similarity and the second similarity
Neighbour user:
According to the similarity factor of the first similarity and the second similarity calculation user: In above-mentioned formula, YW (ui,uj) indicate user uiWith user uj's
Similarity factor;The similarity factor is bigger, illustrates that the similitude between user is higher;
Using the high user of similarity as the neighbour user of user to be recommended;
This preferred embodiment neighbour user selection unit determines the neighbour user of user to be recommended by similarity factor, is
Subsequent music recommendation unit carries out personalized recommendation and lays a good foundation, specifically, the first similarity considers being total between user
With scoring music, the second similarity considers scoring of the user to all music, comprehensive first similarity of the similarity factor and the
Two similarities obtain accurate neighbour user.
Preferably, the music recommendation unit is used to recommend music to user to be recommended according to neighbour user:
According to the scoring of user in neighbour user's set, predict user to be recommended to the recommendation factor for the music that do not score:In above-mentioned formula, H indicates neighbour user's collection of user to be recommended
It closes, pjdIndicate neighbour user ujTo music xdScoring, LGikIndicate user u to be recommendediTo the music x that do not scoredThe recommendation factor;
Recommend music according to sorting from large to small to user to be recommended for the recommendation factor.
This preferred embodiment music recommendation unit is realized by calculating user to be recommended to the recommendation factor for the music that do not score
Personalized accurate recommendation, improves user satisfaction.
It is capable of the intelligent earphone of autonomous control using the present invention, chooses 5 users and test, respectively user 1, user
2, user 3, user 4, user 5 recommend accuracy and user satisfaction to count music, compare compared with intelligent earphone,
What is generated has the beneficial effect that shown in table:
|
Music recommends accuracy to improve |
User satisfaction improves |
User 1 |
29% |
27% |
User 2 |
27% |
26% |
User 3 |
26% |
26% |
User 4 |
25% |
24% |
User 5 |
24% |
22% |
Finally it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than the present invention is protected
The limitation of range is protected, although explaining in detail referring to preferred embodiment to the present invention, those skilled in the art are answered
Work as understanding, it can be with modification or equivalent replacement of the technical solution of the present invention are made, without departing from the reality of technical solution of the present invention
Matter and range.