Computer-implemented system and method for matching agents with callers in an automated call center environment based on user traits
View Patent ↗A computer-implemented system and method for matching agents with callers within a call center environment is provided. A database of traits for agents within a call center is maintained. A call is received into the call center and a voice recording is obtained from a caller during the call. The voice recording from the caller is analyzed by measuring voice characteristics of the voice recording and by identifying traits of the caller based on the voice characteristics. The caller traits are compared with the traits for the agents. One or more of the agents similar to the user are identified based on the trait comparison. One of the similar agents is selected and the call is transferred to the selected agent.
1. A system for matching agents with callers within a call center environment, comprising:
a database of traits for agents within a call center;
a call module to receive a call into the call center and to obtain a voice recording from a caller during the call;
an analysis module to analyze the voice recording from the caller by measuring voice characteristics of the voice recording and by identifying traits of the caller based on the voice characteristics, comprising:
a collection module to generate graphs each comprising multiple overlapping population distribution curves for subtraits of a common trait associated with one of the voice characteristics;
an identification module to identify a single position of the voice characteristic on each of the graphs; and
a selection module to determine the trait for each graph by selecting one of the multiple subtraits as the trait representative of the voice characteristic based on a relationship of the position with each of the population distribution curves for that graph;
a comparison module to compare the caller traits with the traits for the agents;
an identification module to identify one or more of the agents similar to the user based on the trait comparison; and
a selection module to select one of the similar agents and transferring the call to the selected agent.
2. A system according to claim 1 , further comprising:
a weight module to assign a weight to particular traits based on a preference of those traits and to use the weights during the selection of the similar agents.
3. A system according to claim 1 , further comprising:
a trait determination module to determine the caller traits by comparing the voice recording with voice samples for populations classified as having that trait and by selecting the caller trait associated with the population having a voice sample most similar to the voice recording.
4. A system according to claim 1 , further comprising:
a trait identification module to identify at least a portion of the caller traits based on information provided by the caller comprising at least one of telephone number, address, and zip code.
5. A system according to claim 1 , further comprising:
a persuasion module to identify a stage of persuasion of the caller, during the call, based on further identified traits in response to an offer provided by the agent and to persuade the caller to accept the offer by providing guidance to the agent regarding further information about the offer based on the stage of persuasion.
6. A system according to claim 1 , further comprising:
an identity confirmation module to confirm an identity of the agent, comprising:
a voice print for the agent that is associated with a pass code;
a receipt module to receive a voice recording of the pass code from the agent during an attempt to sign in to the call center;
a voice print module to generate a voice print based on the received voice recording by identifying one or more voice characteristics in the voice recording, by representing each voice characteristic as a graph, and by combining the graphs of the voice characteristics as the voice print for the agent;
a voice print comparison module to compare the generated voice print with the stored voice print for the agent; and
a score module to assign a similarity score to the generated voice print based on the comparison and to confirm the identity of the agent when the score satisfies a similarity threshold.
7. A system according to claim 1 , wherein the voice characteristics comprise one or more of pitch, range of resonance frequencies, average frequencies associated with particular vowels, spectral tilt, breathiness, creakiness, nasality, speed, pausing, and measured and derived intonation features.
8. A system according to claim 1 , wherein the similar agent is selected based on at least one of a number of traits shared with the caller, availability of the agent, and demand for agents in the call center.
9. A system according to claim 1 , wherein the caller traits and the agent traits each comprise at least one of origin, accent, intelligence level, education, age, social characteristics, emotion, appeal, personality, speech rate, gender, competence, and comprehensibility.
10. A computer-implemented method for matching agents with callers within a call center environment, comprising steps of:
maintaining a database of traits for agents within a call center;
receiving a call into the call center and obtaining a voice recording from a caller during the call;
analyzing the voice recording from the caller, comprising:
generating graphs each comprising multiple overlapping population distribution curves for subtraits of a common trait associated with one of the voice characteristics;
identifying a single position of the voice characteristic on each of the graphs; and
determining the trait for each graph by selecting one of the multiple subtraits as the trait representative of the voice characteristic based on a relationship of the position with each of the population distribution curves for that graph;
comparing the caller traits with the traits for the agents;
identifying one or more of the agents similar to the user based on the trait comparison; and
selecting one of the similar agents and transferring the call to the selected agent,
wherein the steps are performed by a suitably-programmed computer.
11. A method according to claim 10 , further comprising:
assigning a weight to particular traits based on a preference of those traits; and
using the weights during the selection of the similar agents.
12. A method according to claim 10 , further comprising:
determining the caller traits comprising comparing the voice recording with voice samples for populations classified as having that trait; and
selecting the caller trait associated with the population having a voice sample most similar to the voice recording.
13. A method according to claim 10 , further comprising:
identifying at least a portion of the caller traits based on information provided by the caller comprising at least one of telephone number, address, and zip code.
14. A method according to claim 10 , further comprising:
during the call, identifying a stage of persuasion of the caller based on further identified traits in response to an offer provided by the agent; and
persuading the caller to accept the offer by providing guidance to the agent regarding further information about the offer based on the stage of persuasion.
15. A method according to claim 10 , further comprising:
confirming an identity of the agent, comprising:
storing a voice print for the agent that is associated with a pass code;
receiving a voice recording of the pass code from the agent during an attempt to sign in to the call center;
generating a voice print based on the received voice recording, comprising:
identifying one or more voice characteristics in the voice recording;
representing each voice characteristic as a graph; and
combining the graphs of the voice characteristics as the voice print for the agent;
comparing the generated voice print with the stored voice print for the agent;
assigning a similarity score to the generated voice print based on the comparison; and
confirming the identity of the agent when the score satisfies a similarity threshold.
16. A method according to claim 10 , wherein the voice characteristics comprise one or more of pitch, range of resonance frequencies, average frequencies associated with particular vowels, spectral tilt, breathiness, creakiness, nasality, speed, pausing, and measured and derived intonation features.
17. A method according to claim 10 , wherein the similar agent is selected based on at least one of a number of traits shared with the caller, availability of the agent, and demand for agents in the call center.
18. A method according to claim 10 , wherein the caller traits and the agent traits each comprise at least one of origin, accent, intelligence level, education, age, social characteristics, emotion, appeal, personality, speech rate, gender, competence, and comprehensibility.