Summary of the invention
Technical problem to be solved by this invention is to provide a kind of television program recommending device and method, and the television programming sequence that will obtain from electronic program guides is liked according to spectators and to be recommended spectators after rearranging.
The invention provides a kind of television program recommending device, comprising:
The epg table input unit is in order to obtain original epg table;
Watch-dog is used for the attribute and the time of program that spectators watch are carried out record;
Spectators like statistic device, are used for the record according to described watch-dog, and dynamic calculation spectators watch the time the TV programme of different attribute;
Programs recommended generating apparatus is used for the result of calculation according to described statistic device, to described original electron program guide rearrangement of obtaining, generates the epg table of recommendation for spectators;
Programs recommended guide tab output device is used for the output of the epg table after the rearrangement.
The present invention and then a kind of TV programme suggesting method is provided comprises the steps:
Obtain the original electron program guide;
The attribute and the time of the TV programme that the monitoring spectators watch;
The statistics spectators are to the viewing time of the TV programme of different attribute;
According to the viewing time of spectators to the TV programme of different attribute, the original electron program guide is resequenced, generate programs recommended guide tab;
Export described programs recommended guide tab.
Described statistics spectators comprise the steps: the time of the watching step of the TV programme of different attribute
Calculate spectators and watch the relative time of different attribute TV programme;
Wherein the relative time of the TV programme of particular community averages respectively;
The average relative time of the TV programme of the different attribute that spectators are watched sorts;
Generate the time statistical form that spectators watch the different attribute program by the order after the comparison.
The present invention is by watching the objective statistics and the dynamic calculation of program to spectators, with original not targetedly epg table resequence, can also generate different programs recommended guide tabs at each kinsfolk, make things convenient for the selection of spectators TV programme.
Embodiment
As shown in Figure 1, for television program recommending device embodiment schematic diagram of the present invention, comprise epgtable input unit 101, watch-dog 102, spectators likestatistic device 103, programs recommended generatingapparatus 104, programs recommended guidetab output device 105.
Wherein, epgtable input unit 101 is in order to obtain original epg table; Watch-dog 102 is used for when spectators carry out remote control by 107 pairs of television sets ofremote controller 106, and the attribute and the time of program that spectators watch are carried out record; Spectators likestatistic device 103, are used for the record according to described watch-dog, and dynamic calculation spectators watch the time the TV programme of different attribute; Programs recommended generatingapparatus 104 is used for the result of calculation according to described statistic device, to described original electron program guide rearrangement of obtaining, generates the epg table of recommendation for spectators; Programs recommended guidetab output device 105 is used for the output of the epg table after the rearrangement.
The present invention can also carry out concrete differentiation with spectators, for niche audience generates pointed programs recommended guide tab, for this reason, the present invention can also comprise that a viewer identity determines device, be used for by obtaining and identifying currently, distinguish different spectators just at the viewer identity of TV reception.
If the user does not need spectators are distinguished, so, the recommendation tables that is generated just can be considered as providing at whole family rating hobby.
In addition, described watch-dog 102 can also be further carries out record to the program channel of program that spectators watch; Described spectators likestatistic device 103, can also further calculate the watch frequency of spectators to different program channels; Described programs recommended generatingapparatus 104, can also be further according to the frequency of watching of the different program channels of described calculating, to described original electron program guide rearrangement of obtaining, generate the epg table of recommendation for spectators.
As shown in Figure 2, for TV programme suggesting method embodiment flow chart of the present invention, at first, obtain original electron program guide (step 201); The attribute and the time (step 202) of the TV programme that the monitoring spectators watch; The statistics spectators are to the viewing time (step 203) of the TV programme of different attribute; According to the viewing time of spectators to the TV programme of different attribute, the original electron program guide is resequenced, generate programs recommended guide tab (step 204); Export described programs recommended guide tab (step 205).
Equally, the present invention can also be current just at the viewer identity of TV reception, to distinguish different niche audiences by obtaining and identifying.
Wherein, described monitoring step can also further be monitored the program channel of program that spectators watch; Described statistic procedure is further added up the watch frequency of spectators to different program channels; The programs recommended guide tab step of described generation further according to the frequency of watching of described different program channels, to described original electron program guide rearrangement of obtaining, generates the epg table of recommendation for spectators.
The present invention adds up spectators to the fancy grade of certain type TV programme by programme attribute, and this kind statistics can realize by setting up database table.Simultaneously because kinsfolk's hobby difference will be set up the tables of data of a correspondence at each kinsfolk in database, so just can be purposive with program commending to each corresponding kinsfolk.
For the statistic processes of database table, the program that can at first watch niche audience by watch-dog is monitored, and calculates the relative time length of different attribute TV programme; Then the relative time length of the TV programme of pairing particular community is carried out consecutive mean; Again average relative time is compared ordering, by the niche audience tables of data in the sequential organization spectators' preference database after relatively sorting.The niche audience tables of data of TV Guide in spectators' preference database sorted the program sequencing table after just can obtaining recommending to sort through recommendation process.
Suppose when niche audience A passes throughremote control 107 RTVremote television 106, watch-dog 102 will be monitored the program that A watched, watched the program more than several minutes to like in the statistic device 103 (it can be spectators' preference database) to A and carried out storage spectators, storage mode will watch the relative time length of different attribute TV program time to add up according to spectators, will dynamically average then at every turn.
Term " relative time length " refers to that spectators watch the ratio of total reproduction time of real time of a certain attribute TV programme and this attribute TV programme.For example, the time that news is play is 30 minutes, and spectators have watched 30 minutes, and spectators' relative time length of watching is 1 (maximum is 1) so; The reproduction time of TV play is 45 minutes, and spectators have equally also watched 30 minutes, and his the relative time length of watching is 2/3 so.
By writing down the relative time length 11 of every kind of attribute program, l2 ... .li, dynamically carry out arithmetic average then:
The average d of different attribute TV programme is stored in the tables of data of spectators' preference database.Big minispread according to arithmetic mean d is exactly the fancy grades of spectators to every kind of attribute TV programme.Such as, the relative time length of news is 1,0.7,0.6 etc., the relative time length of TV play is 0.4,0.6,0.8 etc., the relative time length mean value of news is 0.8 so, the relative time length mean value of TV play is 0.6, because 0.8>0.6, illustrate that these spectators prefer watching news program, long-term data accumulation will be determined the favorable rating of spectators to the specific program attribute more.
Program in the electronic program guides has obtained the program after recommendation is sorted after the hobby tables of data of niche audience is passed through recommendation process, this program guide will be presented in the television image.In the program sequencing table after recommendation, can have multiple at putting in order of the concrete program channel of every kind of attribute, the program channel that a kind of preferred embodiment is often watched for spectators is presented at topmost, that is to say, the program of certain generic attribute is watched that the time is long more, and then the program location of this generic attribute is forward more; To program ordering with generic attribute, can sort according to the frequency of watching to different channel, watch that frequency is high more, then the position is forward more.Certainly, other arrangement mode is suitable for too.
Spectators likestatistic device 103 and can be realized by spectators' preference database, and as shown in Figure 3, the tables of data for the spectators of spectators' preference database correspondence wherein mainly comprises three fields: user ID, programme attribute and average d.Wherein user ID has uniqueness, and each spectators will obtain a value; Programme attribute is divided in electronic program guides, as news, amusement, physical culture etc.; Dynamic calculation when average d watches TV programme according to niche audience.
Fig. 4 is the flow implementation illustration intention of the programs recommended guide tab of generation of the present invention, determine user's ID (step 401), in database, inquire about corresponding tables of data (step 402) then, read original epg table (step 403) then, then, tables of data according to the specific user is resequenced (step 404) to epg table, the program guide after obtaining recommending to sort.The algorithm of ordering has many kinds, such as bubble sort, quicksort, Shell sorting, heapsort etc.
Fig. 5 is the embodiment of the TV Guide table after being presented on spectators' a kind of preferable ordering in front after the recommendation ordering, and the attribute section of italics represents that this listing is to arrange according to the hobby of niche audience.And in specific object such as news, putting in order of program channel is to watch the height of the frequency of this program channel to arrange at ordinary times according to spectators.This embodiment shows that in attribute news secondly the news program that spectators watch CCTV-1 through regular meeting often watches news program of SXTV-1 or the like.The spectators' programme attribute and favorite program channel that will go earlier to select oneself to like watched like this, and the program channel of next selecting to like for own time goes to watch.
Fig. 6 is presented on the embodiment of a kind of randomly ordered TV Guide table in face of the spectators after recommending ordering.Wherein at the program channel of specific object (as news) adopt be at random, a kind of arrangement mode arbitrarily.