EP1955539A2

EP1955539A2 - Optimally selecting tv programs with time margins

Info

Publication number: EP1955539A2
Application number: EP06821467A
Authority: EP
Inventors: Wim Verhaegh; Gerrit Hollemans; Johannes Henricus Maria Korst
Original assignee: Koninklijke Philips Electronics NV
Current assignee: Arris Global Ltd
Priority date: 2005-11-21
Filing date: 2006-11-15
Publication date: 2008-08-13
Also published as: WO2007057852A2; WO2007057852A3; JP2009516952A; RU2008125166A; CN101313566A

Abstract

A system (800), apparatus (700), and method are provided to select a best set of TV programs for reception and recording by at least one tuner, said TV programs being broadcast over a set of channels during a given time period [b, e] having a beginning time b and an ending time e. Partial and entire TV programs are considered and for each partial or entire TV program a user preference value is provided for its reception and recording. Further, pre-rolls and post-rolls are included in order to insure that partial TV programs are entirely received. A minimum-cost network algorithm is used that solves the selection and viewing problem but is not necessarily optimal. Additional time is allowed, before or after reception of a part of a TV program, to ensure that the entire part of the TV program has been received.

Description

OPTIMALLY SELECTING TV PROGRAMS WITH TIME MARGINS

The present invention relates to a system, apparatus and method that uses time margins to make (near) optimal selections of partial TV programs to be recorded using a limited number of tuners during a given time interval and distinguishes between switching to another TV channel after recording a first program and staying tuned to the same channel by enforcing a time margin in the former situation. A part is defined as some or all of a TV program.

Personal video recorders (PVRS) are available to record TV programs, either on hard disk or on DVD. Programming these PVRs is usually accomplished using an electronic program guide (EPG), by simply clicking on the TV programs to be recorded. Additionally, a recommender may be available that can predict, for an upcoming TV program, how much a viewer will like it.

With the increasingly available television channels via terrestrial, satellite, or cable connections, the task for the viewer to select TV programs to view and record is rapidly becoming too large to be handled manually. With the advent of digital television, the number of channels, and hence the number of options, to choose from is becoming even larger. Therefore, the viewer is no longer able to overview all available programming content and it becomes more likely that the viewer will miss TV programs that would be of interest to the viewer.

Because of the large number of TV programs available each week, printed TV program guides are cumbersome and electronic program guides (EPGs) have been developed as a solution to this problem. EPGs present the available TV programs for a number of channels on a TV screen. However, only a limited number of TV programs can be shown at one time on the TV screen, i.e., the number of channels and the length of the time interval of the portion shown on the screen is very limited. This results from the poor resolution of a TV screen for presentation of this kind of textual information. And, even if the resolution were high enough, the task of selecting from the very large amount of available TV content would overwhelm the average viewer. In order ameliorate the viewer's problem of what partial TV programs to select and when and how to arrange viewing time among various TV channels, i.e., the problem of developing a viewer-specific viewing schedule that ensures that a selected partial TV program is completely received, EPGs offer an option of searching (filtering) by keyword(s) but do not allow time between selections to complete receipt thereof. In this way EPGs can reduce the number of TV programs to be screened and selected by the viewer to a manageable number but cannot handle a lack of completion of a partial TV program caused by switching channels. Another way for the viewer to cull a plurality of uninteresting-to-the-viewer TV programs is through the use of recommender systems based on artificial intelligence (AI) technology but the same lack of completion problem is present. These AI recommender systems maintain a preference profile of the viewer, indicating what the viewer likes and dislikes, and employ this profile to score each newly offered TV program as to what extent the viewer will like the newly offered TV program. Then, the TV programs scoring in excess of a pre -determined viewer-specific tolerance are highlighted in an EPG or the viewer is provided a list of these TV programs. A problem with these solutions, however, is that they do not take into account whether or not TV programs overlap in time, so it remains for the viewer to compile a viewer-specific schedule of top-scoring partial TV programs to receive and to allow enough time between them for completion of receipt of a selected part. In other words, whereas filters and recommenders come up with a list of individual TV programs, such a list is an incomplete solution to a viewer's problem of creating a schedule of TV programs to watch, especially there is no way to leave enough time between partial TV programs selected to allow completion when channels are to be switched between receptions thereof.

A similar situation exists for recording TV programs.

A solution is needed for efficient selection of partial TV programs for a given time interval that allows time for completion when channels must be switched and that maximizes predetermined viewer preferences when the viewer has only a finite number M > 1 of tuners available with which to receive/record selected partial TV programs.

The system, apparatus, and method of the present invention provide an efficient algorithm that is guaranteed to make (near) optimal selections of partial TV programs including time margins when channels are switched between successive partial TV programs and with a limited number of tuners, i.e., a selection that nearly maximizes a sum of the viewer preference values associated with the selected partial TV programs while guaranteeing complete reception. It is assumed that these preferences are known to, e.g., a recommender that provides a first filtering of available partial TV programs before an (near) optimal selection is made using the present invention.

Given a time interval in which a viewer wants to select partial TV programs, all partial TV programs that fall in this interval are first collected and a value function for each collected partial TV program is obtained. Second, the begin and end time points of each partial TV program are determined as nodes of a directed graph G and a set S is determined of partial TV programs having at least one of a begin and end time point within the given time interval. Third, a time 0 source and a time ∞ sink node are created and a graph of G is drawn from the source through the successive time-ordered nodes of G to the sink, i.e., a central timeline or chain of edges is created from the source node through each successive time-ordered node of G, wherein, each edge has a cost of '0' and a capacity of M, where M > 1 is the number of tuners available to receive a TV program.

Next, a similar timeline for channel c is constructed for the partial TV programs in S that are broadcast on each channel c, in order to distinguish between a TV channel that is already being received and one that is not being received at a given time. That is, in addition to the nodes already included in the central timeline, each TV program in S that is broadcast on TV channel c is further divided into zero or more partial TV programs represented by begin and end node pairs on the central timeline and having edges therebetween not on the central timeline but on a timeline for the channel c such that it is possible for every TV program only to start being received at its begin time or at the end time of another TV program, and such that every TV program stops being received at its end time or at the begin time of another TV program. Every such resulting partial TV program is added to S and has an associated viewer preference value determined by a viewer-specific value function. As shown below in the detailed description, no loss of optimality results from this restriction (see Theorem 1 below).

Key in the present invention is thus the graph G comprising the central timeline, from and to which run the whole and partial TV program edges included, for each TV channel c, in a timeline for channel c, which edges represent the negative preference value of these whole and partial parts of the TV programs and which whole and partial parts are contained in the set S. Finally, a time margin is included in the model, i.e., the graph G, as a time difference on edges running from the central timeline to a timeline for a channel, and this time margin eliminates parallel edges within a timeline for a channel and is needed for properly representing viewer preference values in the timeline for the channel.

Given this directed graph G, a minimum-cost network algorithm is applied thereto that determines the subset of partial TV programs S', for the given time interval, that provides near maximum viewer satisfaction, i.e., nearly maximizes the total of viewer preference values for the set S' of partial TV programs. The selected subset is guaranteed to be nearly optimal and the algorithm selecting the subset is guaranteed to run in polynomial time.

FIG. 1 illustrates a number of TV programs offered over time t;

FIG. 2 illustrates a directed graph for the TV programs of FIG. 1 ;

FIG. 3 illustrates an extended graph for the example of FIG. 1 that includes partial TV programs;

FIG. 4 illustrates an extended graph for the example of FIG. 1 that includes partial TV programs having value functions that are linear with penalties and costs of related edges for TV program A;

FIG. 5 illustrates a graph for the example of FIG. 1 for the case with pre-rolls and post-rolls of five minutes only showing the shows of the first channel and the central timeline;

FIG. 6 illustrates a graph for the example of FIG. 1 for the case with pre-rolls and post-rolls, when only complete TV programs are to be received;

FIG. 7 illustrates an apparatus for performing the method of optimal selection of partial TV programs according to the present invention; and

FIG. 8 illustrates a system for receiving, recording and displaying a set of partial TV programs selected in accordance with the present invention.

The system, apparatus, and method of the present invention provide a way of determining an optimal TV experience for a viewer for a given time interval, the experience comprising watching as well as recording partial preferred TV programs using at least one tuner. The system, apparatus, and method of the present invention provide a mathematical formulation and solution strategy for obtaining an optimal solution to a viewer's TV program scheduling problem. The scope of the system, apparatus, and method of the present invention includes the problem of what TV programs to select for watching or recording with a given number of available tuners based on a pre-determined viewer preference for all available TV programs being broadcast in given time period.

A simplified problem setting is described first followed by an extension of the simplified problem setting to include more factors and to elaborate the algorithms developed for the simplified problem, in order to solve more complex problem settings.

In a preferred embodiment of the system, apparatus, and method of the present invention, several assumptions are first made:

1. a preference value or score is provided for each of a plurality of candidate TV programs, indicating how much a viewer prefers the candidate TV program;

2. each preference score is given explicitly by the viewer or by a recommender tool;

3. the total preference score for a number of TV programs being selected is given by a linear sum of the provided preference scores of the individual TV programs; and

4. TV programs that are selected for watching are assumed to be watched at the very moment they are broadcast; if a TV program is watched later, it is simply assumed to be recorded during its broadcast.

Each of the simplified and extension thereof are modeled as a directed graph to transform the problem into a minimum-cost network flow problem. Since minimum-cost network flow problems can be solved in polynomial time, the corresponding TV program selection problem can be solved efficiently.

The first model selects complete TV programs using one tuner. The problem of selecting complete TV programs is defined as follows.

Definition 1 (Program selection problem (PSP)). A set S of TV programs s is given and for each TV program s a begin time b_s and end time e_s is also specified. Furthermore, for each TV program a value v_s > 0 is given, which reflects the total value if this TV program is watched (completely). The question is to determine a subset S c S such that the TV programs in S do not conflict in time, i.e.,

for all s, t e S , s ≠ t, and such that the total value of this subset, given by

SG S' is maximized.

The value v_s of a TV program can be given by the viewer explicitly, or it can be estimated using a recommender. Considering the recommender score p_s of a TV program s as a preference density, the total value is given by v_s = p_s(e_s - b_s). Maximizing the total value over a number of TV programs in a certain time interval then corresponds to maximizing the average preference density.

The TV program selection problem can be formulated as a minimum-cost flow problem as follows. The set V of nodes of a network G is given by a source node and a sink node, and a node for each begin or end time of a TV program. So,

V = {0,∞} u U {b_s,e_s},

where the node 0 represents the source and ∞ represents the sink. Denoting the nodes (i.e., times) in V by ti, ... , t_n the n distinct times can then be arranged in increasing order. Here, it is assumed that all times are larger than the source time ti = 0, and smaller than the sink time t_n = oo.

The set E of edges consists of two parts. First, there is an edge on a timeline from source to sink that includes the node edge (t_l5 t₁₊i) for all i = l,...,n-l, having zero cost, by definition. Secondly, for each TV program s ε S, an edge (b_s, e_s) is added, with cost -v_s. All edges are assumed to have capacity one.

For an example of the graph construction, consider the set of TV programs as indicated in FIG. 1, covering TV programs A, B,..., J in a time interval from 18:00h to 22:00h. The corresponding set V of time points is given by

F= {0, 18:00, 19:00, 19:30, 20:00, 20:30, 21 :00, 22:00, ∞}, and the set of edges is as given in FIG. 2.

Now, determining a non-overlapping subset S with maximal value corresponds to finding a flow of one in G from the source to the sink with minimum cost. If in this minimum-cost flow an edge corresponding to a TV program carries a flow of one, then this TV program can be included in S ; otherwise not. Hence, the problem can be solved by means of a minimum-cost flow algorithm. In other words, PSP can be solved efficiently.

As in the above formulation all capacities are one, the minimum-cost flow problem reduces to a shortest-path problem (from source to sink), where the (negative) costs are used as edge lengths.

The min-cost flow problem: Given a directed graph G = (V, E). Let there be a cost C_1J and capacity U_y = 1, associated with each edge ij in E, and a demand (supply) by associated with each node v in V. The problem is to find a flow of minimum cost which satisfies the supply/demand constraints for each node,

and the edge constraints,

o ≤ Λ ≤ v

The cost of the flow is defined as

In the present invention, the capacity of each edge is 1, by definition. Furthermore, bi = -1, b_n = 1, and bj = 0 for i = 2,...,n-l.

Multiple TV programs can be received in parallel whenever more than one tuner is available, and the additionally selected TV programs are recorded. The problem is also relevant in a two-stage TV programming approach, where one first determines what TV programs to receive, which are then stored in a persistent memory, and then determines a TV schedule from all content that has been recorded in the persistent memory. Definition 2 (Multi-tuner program selection problem (MPSP)). Given is a set S of TV programs and for each TV program s a begin time b_s, an end time e_s, and a value v_s are given. Furthermore, a pre-determined number M of tuners is also given. The question is to determine a subset S c S such that at all times at most M TV programs have been selected, i.e., for all times x

{y e S'\ b_s ≤ x < e_s } ≤ M,

and such that the total value of this subset, given by V v₅ is maximized.

MPSP is solved in a similar way as PSP, using an equivalent network flow transformation as presented above. The only difference is that for MPSP the capacities of the edges (t_utn-i) are changed into M to reflect the availability of M tuners; the capacity of the TV program edges remains one, as it is not desirable for a program to be received multiple times.

Next, a minimum-cost flow of M from source to sink is determined, and again each TV program for which the corresponding edge in the graph carries a flow (of one) in this found minimum-cost flow solution is included in S .

Next, the above-described single tuner solution for complete TV programs is extended to include watching some TV programs only partially. To this end, a value function v_s(l, r) is introduced for each TV program s that indicates the value for watching the program during the partially open interval [1, r), with b₅ < 1 < r < e₅. For this value function, it is assumed that

v_s (b₅, e_s) = v_s, v_s(li, ri) > V₅(I₂ , r₂) if [l_h ri) => [I₂ ,r₂), and v_s(x, x)<0.

Definition 3 (Partial program selection problem (PPSP)). Given is a set S of TV programs, and for each TV program s a begin time b₅ and end time e₅. Furthermore, for each TV program a value function v₅ is given. The question is to determine for each TV program s ε S a (possibly empty) set I₅ = { [Ii , ri), ... , [l_k ,r_k )} ) of intervals during which to watch it, with b₅ < Ii < ri < I₂ < r₂ < • • • < l_k < r_k < e₅, such that the selected intervals of the TV programs do not conflict in time. This means that for all s, t e S', s ≠ t, and for all [/, r) e I_s , [l', r') e /₍ we have l ≥ r' v I' ≥ r , i.e., any interval of s is either later than any interval of t, or vice versa. Note that it is allowed that one interval starts exactly at the same moment that another one ends, as we used half-open intervals. The objective is that the total value of the selected intervals, given by

∑ ∑v.Q.r), s≡S [l,r)el_s

is maximized.

Although the value function can be of almost any form, in the following discussion it is restricted to convex value functions, which are discussed in the next section. An even more restricted form, which can be handled more efficiently, is a linear value function with penalties, also discussed below.

Convex value functions

Convex value functions are value functions v that satisfy the condition that for all α e [0,1] the following holds

V(OcZ₁ + (1 - oc )/₂ , OCr₁ + (l - oc )r₂) < αv(Z₁,r₁ )+ (l -α)v(Z₂,r₂ ) .

Most practical value functions will satisfy this condition. Example convex value functions are linear functions such as

or quadratic functions such as

One of the properties of a convex value function is that it allows reduction of the number of possible points at which to switch from one TV program to another, as the next theorem states.

Theorem 1. If for all TV programs s e S, v_s(l, r) is convex, then, without loss of optimally, PPSP can be restricted to solutions with for all s e S and [1, r) e I_s

In other words, a TV program only starts being received at its begin time or at the end time of another TV program, and it only stops being received at its end time or at the begin time of another TV program.

Proof. Consider an interval [1, r) e I _s of a TV program s e S that does not satisfy

(1). Just before time 1, the maximum number of tuners must have been used, as otherwise it would be possible to lower 1 and improve the solution. As reception of TV program s occurs from time 1 onwards, there must be a TV program t ε S that stops being received at time 1, i.e., for which I_t contains an interval [1', r') with r' = 1 . Because 1 does not satisfy (1), r' < e_t. Now consider two alternative solutions, one with 1 replaced by Ii = 1 - ε and r' replaced byr/ = r' -ε = l_λ , and one with 1 replaced by h = 1 + ε and r' replaced by Z₁ = r' + ε = I₁ . Compared to the original solution, the cost of the first solution differs an amount

Δ₁ = v_J(/ -ε, r) -v_J(/, r) + v,(/',r' -ε ) -v,(/',r')

from the value of the original solution, and the cost of the second one differs by an amount

Δ₂ = v_J(/ + ε, r) - v_J(/, r) + v,(/',r' + ε) -v,(/',r') .

Therefore, it follows that A₁ +A₂ _ v_s(l -ε,r) + v_s(l + ε,r)

-v_s ⁽J,r)

_{+ Vl}(i^>y-e)_{+ Vl}(i^>y₊e) __Vt(r^

> 0 + 0 = 0 where the latter inequality is based on the convexity property of v_s and v_t. Hence, at least one of the differences A₁ and A₂ is non-negative, i.e., at least one of the two alternative solutions is at least as good as the original one. So, without loss of optimality, one can either increase 1 and r' until they become e_t or r (in the latter case the interval [1, r] can be removed from I_s), or one can decrease 1 and r' until they become b_s or 1' (in the latter case [1', r') can be removed from I_t).

In a similar way one can prove that (2) can be assumed without loss of optimality.

As a result of Theorem 1, it is only necessary to consider a limited number of time points at which to start and stop watching a TV program partially, provided the value functions are convex. Based on this, the minimum-cost network of FIG. 2 can be extended as shown in FIG. 3. The costs for the TV program edges are determined by the value for the corresponding part of the TV program. An optimal flow in the extended graph now corresponds to an optimal solution of PPSP. In such a flow, multiple edges per program may be used, which corresponds to receiving the TV program during multiple intervals. For instance, an optimal flow might run through edges Ii and I₇ in the extended graph, meaning that TV program I is selected from 19:00-19:30 and from 20:00-21 :00. As there is only a flow of one, the same time interval is never covered twice. Furthermore, selecting two consecutive parts is at most as good as selecting the combined part (e.g. choosing I₅ is at least as good as choosing Ii and 12), hence adjacent intervals can be merged (and are automatically merged if that is strictly better).

Although the number of edges in the extended graph is much larger than in the original graph, it is still polynomially bounded. More precisely, the number of edges per TV program is quadratic in the number of different time points. Furthermore, since there is only one tuner, the number of edges can be limited between each pair of time points to only the best one, thereby limiting the total number of edges to O(n ).

Linear value functions with penalties In this alternative, the form of the value functions is restricted even further, to linear functions with penalties for finishing a TV program early or starting a TV program late. More formally, consider value functions of the form

where

0 if = b

>:' (/) = c? if / > b.

are functions that give a penalty if TV program s is finished early or started late, respectively. Note that this value function obeys the assumptions and that it is convex, as all three terms in the right-hand side of (3) are convex (i.e., the penalty functions are concave). Given a value function as in (3), we can define another extended minimum-cost network, as shown in FIG. 4. More formally, this graph is constructed as follows.

The set V of nodes first of all contains a node for each time point ti,...,t_n. Next, for each TV program with time points b_s = t_l5 and e_s = t, with j - i > 1, an extra node f_k can be constructed for each intermediate time point, i.e., for each k = i+1, ... ,j - 1. In FIG. 4, the node labeled "19:00A" is an example of such a node for an intermediate time point.

The set of edges, all with capacity one, is defined as follows. First, there is an edge (t_utn-i) for all i = 1, ... , n-1, as before with cost zero. Secondly, consider each TV program se S, with time points b_s = I₁ and e_s = t,, and one again distinguishes two cases, as follows.

• If j - i = 1, then add an edge (b_s, e_s) with cost -v_s.

• If j - i > 1, then add a number of edges: t,^ - 1,

(i) an edge (t, , t°_+i ) with cost - v_s e - b

(ii) an edge (f_k , f_k+l ) for each k = i+ 1,..., j - 2 with cost - v_s ^{k+l k} e - b. (iii) an edge (f ,t ) with cost - v_s — — — ,

(iv) an edge (f_k , t_k ) for each k = i + 1 , ... , j - 1 with cost cf , and (v) an edge (t_k,f_k) for each k = i + l,..., j - 1 with cost cf .

Now PPSP with linear value functions with penalties corresponds to finding a minimum- cost flow of one from source to sink in this graph. Again, multiple parts per TV program may be selected, as with the graph of FIG. 3.

For the correspondence between a flow in the graph and a selection of partial TV programs, consider again the example in FIG. 4. TV program A is completely selected if both edge (18:00,19:00A) and edge (19:00A, 19:30) carry a flow of one. In this situation the edges (19:00,19:00A) and (19:00A,19:00) will not carry a flow. The cost of the flow over these former two edges exactly add up to -V_A, corresponding to a value V_A for the selection of TV program A. Only selecting the first part of A corresponds to a flow of one through edges (18:00,19:00A) and (19:00A, 19:00), resulting in a cost -2v_A/3 + c_A° , i.e., a value 2V_A/3 for this first part and a penalty c^ for finishing A early. Only selecting the second part of A corresponds to a flow of one through edges (19:00,19:00A) and (19:00A, 19:30), resulting in a cost -V_A/3 + cj^s , i.e., a value V_A/3 for this second part and a penalty c}^s for starting A late.

The number of nodes and edges in the extended graph is still polynomially bounded in the size of the instance, as the total number of nodes is O(|S|n), and the number of edges is

O(n + |S| + |S|n) = O(|S|n).

A next extension of PSP is to allow multiple tuners as well as partial TV programs, as given by the next definition.

Definition 4 (Multi-tuner partial program selection problem (MPPSP)). Given is a set S of TV programs, and for each program s a begin time b_s and end time e_s. Furthermore, for each program s a value function v_s is given, and a number m of tuners is given. The problem is to determine for each TV program s e S a (possibly empty) set I_s = {[li, ri), ... , [l_k , r_k )} of intervals during which to watch or record it, with b_s < li< r i < I₂ < r₂ < • • • < 4 < r_k < e_s, such that the TV programs can be received in the selected intervals with m tuners, i.e., for all times x we have

and such that the total value of the selected intervals, given by

is maximized.

Unfortunately, a straightforward application of the extended graph for convex value functions with adapted capacities, as discussed above, does not work. The reason for this is that in that extended graph one may select several parts of the same TV program to be received in parallel. For instance, for the example of FIG. 3 with three tuners the resulting solution might be to receive TV program I and partial TV programs Ii and I₅, meaning that the part of TV program I from 19:00-19:30 will be received three times.

For linear value functions with penalties, nevertheless, the graph as shown in FIG. 4 applies. The only modification is again that we have to increase the capacities of the edges (t^tn-i) from 1 into m, and a minimum-cost flow of m from source to sink must be found. The reason this graph can be reused is that for each TV program and each time interval there is at most one edge covering it, so no part is received more than once.

Referring now to FIG. 7, an apparatus (700) is shown for selecting a set S (701.1) from a memory module (701) storing a given set of N > 1 TV programs s, each having a begin time b_s, an end time e_s, and a viewer-preference function v_s , the latter being provided by the viewer or a recommender. For one tuner the value function is selected from the group consisting of convex function and linear with penalties. For more than one tuner, the value function is linear with penalties. The apparatus further comprises a processor module (702) that executes or itself further comprises a minimum cost network algorithm module for selecting an optimal set S' c S including partial TV programs to be received and possibly recorded by M > 1 tuners (703). A preferred embodiment includes pre-roll and post-roll: There may be some inaccuracies in the begin times and end times of programs. So, if a program s is selected for reception, then in order to receive the program completely, a tuner has to tune into the corresponding channel c from time ε_c(b_s) < b_s until time λ_c(e_s) > e_s. Here, ε_c(t) is a function that indicates from how early one should tune into channel c in order to receive the content scheduled from time t onwards, and λ_c(t) indicates until how late one should tune into channel c in order to receive the content scheduled up to time t. Both ε_c and λ_c are assumed to be strictly increasing.

The amount b_s - ε_c (b_s) is called the pre-roll of program s, and the amount λc (e_s) = e_s is called the post-roll of program s. The main effect of the pre-roll and post-roll of programs on the program selection problem is that if two successive programs are to be received with the same tuner, and these programs are on different channels, then there should be some time between the scheduled end time e_s of the first program and the scheduled begin time b_s of the second one if the programs have to be received completely. For two successive programs on the same channel, there is no such issue.

To reflect the fact that a longer interval is needed to obtain the full value of a program s on channel c, the value function given in (3) is redefined as

vΛl,r) = v_s pf (r) - pf (l), (4)

where the penalty functions are also redefined as

Alternatively, the penalty functions can be redefined as if r<e_r _; ifε_c{b_s)≤l≤b_s

if />Λ.

These latter penalty functions reflect that the penalty becomes less the more time is included beyond the end points. As these penalty functions are still concave, the redefined value function of (4) is still convex. So, Theorem 1 can be applied, and the only considerations are adapted begin points ε_c(b_s)and end points λ_c(ejfor starting or stopping reception of a program.

Definition 5 (Program selection problem with pre -rolls and post-rolls (PSPR)). Given is a set C of channels, with for each channel c an earliness function ε_c and lateness function λ_c. Furthermore, a set S of TV programs is given, with for each program s a begin time b_s, an end time e_s, and a (redefined) value function v_s as given by (4). Finally, a number M > 1 of tuners is given.

The question is to determine for each program se Sa (possibly empty) set

of intervals during which to watch or record it, with

ε_Ci (b_s) ≤l₁ <r₁ <l₂ <r₂ <...<l_k <r_k ≤ λ_Ci (e_s)

such that one never has to tune into more than M channels at the same time, i.e., for all times x

[c_s I s e S A [/, r) e I_s A 1 ≤ x < r\< M, and such that the total value of the selected intervals, given by

∑∑v,(/,r),

is maximized.

For PSPR it is assumed that the TV programs on a particular channel do not overlap in time, i.e., [b_s, e_s) Pl [b_t, e_t) = 0 for all s, t e S, s ≠ t, with c_s = c_t. Note that the selected intervals for two such TV programs may overlap, while still being received with the same tuner. For instance, for the example of FIG. 1, assuming a pre-roll and post-roll of five minutes, an interval may be selected [17:55,19:35) for program A and an interval [19:25,20:05) for program B, because they are on the same channel.

Considering this to be a minimum-cost flow problem, due to the pre-rolls and post- rolls, it is not trivial to construct a corresponding graph in which the costs are correctly modeled and in which no programs can be selected multiple times in the same time interval. Therefore, a graph is constructed as shown in FIG. 5, which has the drawback, however, that not all possible solutions to PSPR can be obtained and correspondingly the solutions can only be guaranteed to be near optimal (this is further discussed below). FIG. 5 illustrates a graph for the example of FIG. 1 for the case with pre-rolls and post-rolls of five minutes, only showing the shows of the first channel and the central timeline. In the cemtral timeline, the labels 'b' and 'e' indicate whether a time point represents the begin time or the end time of a TV program, respectively, where an additional subscript indicates the respective TV program. Next, the following are defined:

ε S A C_S = c}, all early begin points on channel c;

T° = {λ_c (e_s ) ε S Λ c_s = c}, all late end points on channel c;

T-^b = {ε_c (&_s ) |s ε S A c_s ≠ c}, all early begin points not on channel c;

T-^e = {λ_c (e_s ) |s ε S A c_s ≠ c}, all late end points not on channel c;

T^h = [J ε 5 }, all early begin points of TV programs in S; and r^e = [J T_c ^e = {λ_c (e_s ε S }, all late end points of TV programs in S.

The set of all time points is given by r = {o,co} u r^b uf . Again, denote the distinct time points in T by ti, ... , t_n with ti = 0, t_n = ∞, and V^t_1+I for all i = 1 ,..., n - 1. The graph again contains a central timeline, consisting of a node for each time point t_l5 i = 1, ... , n, and an edge (t_l5 t₁₊i), i = 1 ,..., n - 1, for each pair of successive time points. The capacity of these edges is set to m, and their costs are zero.

Next, nodes and edges are defined that are associated with each TV channel c e C. As Theorem 1 indicates, tuning into channel c should be started at a time point at which either a program on TV channel c starts or at which a program on another TV channel ends. The former set of time points is given by T^ , and the latter is given by T^ . Similarly, tuning into channel c should be stopped at a time point at which either a program on channel c ends or at which a program on another channel starts. The former set is given by T_c ^e , and the latter one by T_c ^h .

When tuning- in is started at time te T^ u T-^c , reception is certain for the content of TV channel c that is scheduled from time ε ^"' (t) onwards, and when tuning-in is stopped at time te T_c ^e u T-^b this means that reception is certain for the content of TV channel c that is scheduled up to time λ^"1^) . Therefore, a set

T_c = T_c ^h U T;} U \ te T_c ^e u T,^h},

is introduced indicating the time point in the broadcast schedule from which and until which content is certain to be received. Now, numbering the distinct time points in T_c by t_cl,..., t_cn in increasing order, a node is introduced in the graph for each time point t_cl, i = 1,

... , n_c and an edge (t_c^₊i), i = 1, ... ,n_c - 1, of capacity one, for each pair of successive time points in T_c.

Next, we add edges between the central timeline and the timeline of channel c, all having capacity 1. To this end, an edge is drawn from each time point t e T_c u T? on the central timeline to time point ε^~ι(t) on the timeline of channel c. Furthermore, for each time point te T_c ^e u T-^b , an edge is drawn from time point λ^"1 (t) on the timeline of channel c to to time point t on the central timeline. Finally, costs are defined for the edges in the timeline of channel c, and the edges to and from the central timeline. To this end, consider each TV program s on channel c, i.e., with Cs = c. Let t_cl and t_CJ be the begin and end time of program s, i.e., t_∞ = b_s and t_CJ = e_s (note that these time points are indeed in T_c). Two cases can be distinguished, as follows:

• If j - i = 1 , then the edge (t_cl, t_CJ) is given a cost -v_s.

• If j - i > 1 , then

(i) the edge (t_cl, t_C;1+i) is given a cost - v_s

(ii) the edge (t_CJ_i, t_CJ) is given a cost - v_s ^' h⁽e_s) -ε_cφ_sy

(iii) the edges (t_ck, t_C;k+i), k = i + 1 , . . . , j - 2, is given a cost - v_s

(iv) any edge (ti, t_ck) with i< k < j is given a cost - v_s h cf , and

(v) any edge (t_ck, t{) with i< k < j is given a cost - v \- c^fe .

Other edges than the ones above get a cost of 0.

For program A on channel c in FIG. 5, where b_A = t_ci and e_A = t^, examples of the five types of edges are

(i) the edge (18:00,18:50) on TV channel c,

(ii) the edge (19:10,19:30) on TV channel c,

(iii) the edge (18:50,19: 10) on TV channel c,

(iv) the edge (19:05,19: 10) from the central timeline to the timeline for TV channel c, and

(v) the edge (18:50,18:55) from the timeline for TV channel c to the central timeline, respectively.

Now the approach to solve PSPR is to find a minimum-cost flow of m from source to sink in the above-constructed graph. The time points at which a flow leaves the central timeline t_l5 i = 1, ... , n, indicate when tuning into a TV channel starts, and the time points at which the flow returns to the central timeline indicate when tuning into the TV channel stops. Furthermore, the edges on the timeline of TV channel c with a flow of 1 indicate the parts of the programs that are received. Solving the above minimum-cost flow problem gives a solution to PSPR, but there is one catch: not all solutions to PSPR can be achieved. For instance, if there is a pre-roll of three minutes and a post-roll of five minutes, then any interval selected for tuning into the corresponding TV channel is at least eight minutes long. In other words, very small tuning intervals cannot be achieved, as one always gets a pre-roll and post-roll. In practice, this will not be a severe restriction. Therefore, the algorithm will be optimal in practical cases. Furthermore, the presented graph can be refined a bit, thereby relaxing this restriction.

Finally, in case no partial programs are to be received but only complete ones, the constructed graph can be reduced significantly. Then, the set T_c only contains the begin and end time points b_s, e_s of TV programs on that TV channel, and only the case j - i = 1 applies for any TV program. FIG. 6 illustrates the resulting graph for this problem instance, where any edge with a program label s has cost -v_s.

Referring now to FIG. 7, an apparatus (700) is shown for selecting a set S (701.1) from a memory module (701) storing a given set of n > 1 TV programs s, each having a begin time b_s, an end time e_s, and a value function v_s that has been redefined such that a longer interval is needed to obtain the entire TV program (part), the latter being provided by the viewer or a recommender. For one tuner, the value function is selected from the group consisting of convex functions and linear functions with penalties. For more than one tuner, the value function is linear with penalties. The apparatus further comprises a processor module (702) that executes or itself further comprises a minimum cost network algorithm module for selecting an optimal set S' c S including partial TV programs with time margins to be received and possibly recorded by M > 1 tuners (703).

Referring now to FIG. 8, a system (800) comprising the apparatus of FIG. 7 to select an optimal set of TV programs that includes partial TV programs for a given time interval and given viewer preferences, and a TV set wherein an incoming broadcast signal (704) is scanned for TV programs selected by the apparatus of FIG. 7 (700) and selected partial TV programs are recorded by the tuners (703) and displayed on the TV (801) all being controlled/executed by processor (702).

While the preferred embodiments of the present invention have been illustrated and described, it will be understood by those skilled in the art, the system, apparatus and method for selecting a best set of partial TV programs with pre-rolls and post-rolls based on maximizing viewer satisfaction expressed as a value function of viewer preferences, as described herein, are illustrative and various changes and modifications may be made and equivalents may be substituted for elements thereof without departing from the true scope of the present invention. In addition, many modifications may be made to adapt the teachings of the present invention to a particular situation, e.g., various optimization algorithms, without departing from its central scope. Therefore, it is intended that the present invention not be limited to the particular embodiments disclosed as the best mode contemplated for carrying out the present invention, but that the present invention include all embodiments falling within the scope of the appended claims.

Claims

CLAIMS:

1. A method for selecting a set of partial TV programs preferred by a viewer in a given time period [b, e] beginning at time b and ending at time e, comprising the steps of: providing a set C of channels c (701.1) each having an associated earliness function ε_c and lateness function λ_c to modify begin and end times of partial TV programs received thereover to include at least one of a pre-roll and post-roll; specifying a set S (701.1) of a number N > 1 partial TV programs s each having a value function v_s , begin time b_s, end time e_s, and each being received over a provided channel c; providing M > 1 tuners (703) to receive and record a partial TV program s e S; constructing a directed graph G = (V, E) comprising a set of nodes V corresponding to said begin time and end time for each s to be received over a channel c and a set of edges E having associated costs and capacities connecting a central timeline of the begin and end nodes each respectively modified by the associated earliness and lateness function (501) to include a pre-roll and a post-roll, for each channel c, a timeline for the channel c of the begin and end nodes for each s to be received thereover (502), and each node of the central timeline with a corresponding node of at least one timeline for a channel c (503) in accordance with a predetermined rule; and applying a minimum-cost network flow algorithm (702.1) to the directed graph G to determine a subset S' c S such that the value function v_s (1, r) over the subset S' is maximized and the partial TV programs in S' are received by the M > 1 tuners without conflicts during the time period [b, e].

2. The method of claim 1, wherein each of the partial TV programs of the set S that are received on a specific channel c e C does not overlap in time with any other TV program on the same channel c.

3. The method of claim 1, wherein the applying step further comprises the step of the value function computing the linear sum of the value function v_s over the subset S' .

4. The method of claim 1, wherein the specifying step further comprises the step of the viewer or a recommender specifying the set S.

5. The method of claim 1, wherein the value function is a linear function having pre-determined penalties for each of finishing a show early and starting a show late.

6. The method of claim 1, wherein the constructing step further comprises the steps of: arranging the nodes and edges of the central timeline (501) in increasing order from a source node at time 0 through the modified begin time b_s and end time e_s of each s e S to a sink at time ∞; and for each channel c, arranging the nodes and edges of the timeline for the channel c (502) in increasing order through the begin time and end time of each s received over the channel c.

7. The method of claim 6, wherein; each edge of the central timeline has a capacity M > 1 and cost 0; each edge of each channel timeline has a capacity 1 and a cost determined according to a first pre-set function, said second node pair of time points being a time from which and a time until which content for a TV program se S is certain to be received over channel c; and each edge that connects the central timeline and at least one channel timeline has a capacity 1 and connects a node b_s, and each node e_s of the central timeline that has been respectively modified by the earliness function and lateness function to the corresponding time point b_s and e_s of the at least one channel timeline, said edge having a cost determined according to a second pre-set function.

8. The method of claim 7, wherein the applying step further comprises the step of the value function computing the linear sum of the v_s over the subset S' .

9. The method of claim 8, wherein the specifying step further comprises the step of the viewer or a recommender specifying the set S.

10. The method of claim 9, wherein the value function is a convex function having pre-determined penalties for each of finishing a show early and starting a show late.

11. An apparatus (700) for reception and delivery of viewer-preferred TV programs (704) during a given time period [b, e] beginning at time b and ending at time e, comprising: a memory module (701) containing a set S of a number N>1 of viewer-preferred TV programs numbered s = l,...,n each having a begin time b_s, an end time e_s, and a value function v_s ; a number M>1 of tuners (703) to receive (and record) a TV program s of S; and a processor module (702) configured to execute a partial program model formulation module to formulate and store in the memory (701) a partial program model of the set S and generate therefrom a directed graph G of partial TV programs having pre-rolls and post- rolls, and a minimum-cost network algorithm module that accesses the memory to select a subset of the generated graph G of partial TV programs to receive and record, such that the selected subset of partial TV programs can be received by the plurality of tuners without conflicts within the time interval [b, e], and the value function over the subset is maximized.

12. A system (800) for scheduling a TV viewing session of partial TV programs preferred by a viewer, comprising: a TV set (801) for viewing a TV program; and an apparatus (700) according to claim 11 for selecting and receiving a viewer's preferred TV programs and displaying on the TV (801) at least a part of the received TV programs during a predetermined time span [b, e] having a beginning time b and an ending time e.

13. An apparatus (700) for reception and delivery of viewer-preferred TV programs (704) during a given time period [b,e] beginning at time b and ending at time e, comprising: a memory module (701) containing data describing a set S of a number N>1 of viewer-preferred TV programs s each having a begin time b_s, an end time e_s, and a value function v_s ; a set of M>1 tuners (703) to receive and record a partial TV program s of S; and a processor module (702) configured to execute the method of claim 1 using the data describing set S contained in the memory module (701) and the set of tuners.

14. An apparatus (700) for reception and delivery of viewer-preferred partial TV programs (704) during a given time period [b, e] beginning at time b and ending at time e, comprising: a memory module (701) containing data describing a set S of a number N > 1 of viewer-preferred TV programs s each having a begin time b_s, an end time e_s, and a value function v_s ; a set of M > 1 tuners (703) to receive (and record) a partial TV program s of S; and a processor module (702) configured to execute the method of claim 4 using the data describing set S contained in the memory module (701) and the set of tuners.

15. A system (800) for scheduling a TV viewing session of partial programs preferred by a viewer, comprising: a TV set (801) for viewing a TV program; and an apparatus (700) according to claim 11 for selecting and receiving a viewer's preferred partial TV programs and displaying on the TV (801) at least a part of the received TV programs during a predetermined time span [b, e] having a beginning time b and an ending time e.

16. A system (800) for scheduling a TV viewing session of partial programs preferred by a viewer, comprising: a TV set (801) for viewing a TV program; and an apparatus (700) according to claim 13 for selecting and receiving a viewer's preferred partial TV programs and displaying on the TV (801) at least a part of the received TV programs during a predetermined time span [b, e] having a beginning time b and an ending time e.

17. A computer program stored in a memory (701), comprising an executable module to perform the method of claim 1 (702.1-3) and a data module (701.1) including a set C of channels and a set S of TV programs as input to the method of claim 2, said set S comprising a number n≥l of preferred partial TV programs s each having a begin time b_s, an end time e_s, and a value function v_s .

18. A computer program stored in a memory (701), comprising an executable module to perform the method of claim 3 (702.1-3) and a data module (701.1) including a set C of channels and a set S of TV programs as input to the method of claim 3, said set S comprising a number n≥l of preferred partial TV programs s each having a begin time b_s, an end time e_s, and a value function v_s .

19. A computer program stored in a memory (701), comprising an executable module to perform the method of claim 9 (702.1-3) and a data module (701.1) including a set C of channels and a set S of TV programs as input to the method of claim 4, said set S comprising a number n≥l of preferred partial TV programs s each having a begin time b_s, an end time e_s, and a value function v_s .

20. A computer program stored in a memory (701), comprising an executable module to perform the method of claim 9 (702.1-3) and a data module (701.1) including a set C of channels and a set S of TV programs as input to the method of claim 5, said set S comprising a number n≥l of preferred partial TV programs s each having a begin time b_s, an end time e_s, and a value function v .