An Investigation into Guest Movement in the Smart PartyOutlineWhat is the Smart Party?Project MotivationSmart Party Simulation ProgramMetricsKey valuesMobility Models TestedTest ProcedureRound 1 ResultsRound 2: Satisfaction OverviewRound 2: Fairness OverviewTopics for AnalysisMoving Versus Not MovingParty stabilization?Initial room seekingInitial room seeking, cont.Population-based modelsSatisfaction based modelConclusionAcknowledgementsReferencesAn Investigation into Guest Movement in the Smart PartyJason Stoops ([email protected])Faculty advisor: Dr. Peter ReiherOutlineProject IntroductionKey metrics and valuesMobility Models, Methods of TestingResultsAnalysisWhat is the Smart Party?Ubiquitous computing applicationSomeone hosts a gatheringGuests bring wireless-enabled devicesDevices in the same room cooperate to select and supply media to be playedSongs played in a room represent tastes of guests present in that roomProject MotivationAre there ways to move between rooms in the party that can lead to greater satisfaction in terms of music heard?Can we ultimately recommend a room for the user?What other interesting tidbits about the Smart Party can we come up with along the way?Smart Party Simulation ProgramBasis for evaluating mobility models (rules of movement).Real preference data from Last.FM is used.Random subsets of users and songs chosenMany parties with same conditions are run with different subsets to gather statistics about the party.Initial challenge: extend existing simulation to support multiple rooms.MetricsSatisfaction: based on 0-5 “star” ratingRating determined by play countExponential scale: k-star rating = 2k satisfaction0-star rating = 0 satisfaction (song unknown)Fairness: distribution of satisfactionGini Coefficient – usually used for measuring distribution of wealth in a population.In Smart Party, wealth = satisfaction.Ratio between 0 to 1, lower is more fair.Key valuesHistory LengthNumber of previously heard songs the user device will track.Used to evaluate satisfaction with current roomSatisfaction ThresholdUsed as a guide for when guest should consider moving.If average satisfaction over last history-length songs falls below sat-threshold, guest considers moving.Mobility Models TestedNo movementRandom movementThreshold-based random movementThreshold-based to least crowded roomThreshold-based, population weightedThreshold-based, highest satisfactionTest ProcedureRound 1: Broad testing to find good values for history length and satisfaction threshold for each model. (25 iterations)Round 2: In-depth evaluation of model performance using values found above. (150 iterations)Ratio of six guests per room maintainedRound 1 ResultsModel History Length ThresholdNo Movement n/a n/aRandom n/a n/aThreshold Random 4 1Threshold Least Crowded4 1Threshold Random, Population Weighted5 0.5Threshold Highest Satisfaction2 2.25Round 2: Satisfaction Overview18 guests / 3 rooms30 guests / 5 rooms60 guests / 10 rooms90 guests / 15 rooms050100150200250Median Overall Satisfaction25th, 75th quartiles shownNOMOVETHRESHOLD LEAST CROWDEDTHRESHOLD RANDOM POP WEIGHTEDRANDOMTHRESHOLD RANDOMTHRESHOLD HIGHEST SATSatisfactionRound 2: Fairness Overview18 guests / 3 rooms30 guests / 5 rooms60 guests / 10 rooms90 guests / 15 rooms00.050.10.150.20.250.30.350.40.45Median Overall Fairness25th, 75th quartiles shown, lower is betterNOMOVETHRESHOLD LEAST CROWDEDTHRESHOLD RANDOM POP WEIGHTEDRANDOMTHRESHOLD RANDOMTHRESHOLD HIGHEST SATFairnessTopics for AnalysisMoving is better than not movingParty stabilization?Initial room seekingPopulation-based models perform poorlySatisfaction-based model performs wellMoving Versus Not MovingMovement “stirs” party, making previously unavailable songs accessibleSongs users have in common changes with movement, depleted slower.NOMOVE RANDOM020406080100120140160180200Random vs. No Move, Median Overall Satisfaction25th, 75th quartiles shown18 Guests / 3 Rooms30 Guests / 5 Rooms60 Guests / 10 RoomsSatis factio nParty stabilization?Do users find “ideal rooms” and stop moving?No! Some movement is always occurring.Cause: Preferences are not static, they evolve over time.0 5 1 0 1 5 2 0 2 5 3 0 3 500 . 10 . 20 . 30 . 40 . 50 . 60 . 70 . 80 . 9R o o m C h a n g e s p e r G u e s t o v e r T i m eT h r e s h o l d H i g h e s t S a t , 3 0 g u e s t s / 5 r o o m sR o u n dM o v e m e n t P r o b a b i l i t yInitial room seeking90% of guests move after round 1Guests have some information to go on after one song plays.Guests that like the first song in a room likely have other songs in common.0 5 1 0 1 5 2 0 2 5 3 0 3 500 . 10 . 20 . 30 . 40 . 50 . 60 . 70 . 80 . 91R o o m C h a n g e s p e r G u e s t o v e r T i m eT h r e s h o l d H i g h e s t S a t , 6 0 g u e s t s / 1 0 r o o m sR o u n dM o v e m e n t P r o b a b i l i t yInitial room seeking, cont.In satisfaction-based model, peak is in round 2All other models peak in round 1.0 5 1 0 1 5 2 0 2 5 3 0 3 5012345678R o u n d - b y - r o u n d S a t i s f a c t i o n6 0 G u e s t s i n 1 0 R o o m sN O M O V ER A N D O MT H R E S H O L D R A N D O MT H R E S H O L D H I G H E S T S A TR o u n dS a t i s f a c t i o nPopulation-based modelsWorse than choosing a room at random!Weighted model performed better as weighting approached being truly random.However, still better than not moving at all.18 guests / 3 rooms 30 guests / 5 rooms020406080100120140160180200Median Overall Satisfaction25th, 75th quartiles shownNOMOVE THRESHOLD LEAST CROWDEDTHRESHOLD RANDOM POP WEIGHTEDRANDOMSa tis factionSatisfaction based modelInformed movement better than random movement.Greater advantage as more rooms are added.Short history length (two songs) used since history goes “stale”.18 guests / 3 rooms30 guests / 5 rooms60 guests / 10 rooms90 guests / 15 rooms050100150200250Median Overall Satisfaction25th, 75th quartiles shownNOMOVE RANDOM THRESHOLD RANDOMTHRESHOLD HIGHEST SATSatis factionConclusionRoom recommendations are a feasible addition to the Smart Party User Device Application.Recommendations based on songs played are more valuable than those based on room populations.Movement is a key part of the Smart Party.AcknowledgementsAt the UCLA Laboratory for
or
We will never post anything without your permission.
Don't have an account? Sign up