Method and system for analyzing separated voice data of a telephonic communication to determine the gender of the communicant
A method and system for determining the gender of a communicant in a communication is provided. According to the method, at least one aural segment corresponding to at least one word spoken by a communicant is identified. The aural segment is then analyzed by applying a gender detection model to the aural segment, and gender detection data is generated based on the application of the gender detection model.
1. A non-transitory computer readable medium adapted to control a computer and comprising a plurality of code segments for determining the gender of a communicant in a communication, the non-transitory computer readable medium comprising:
a code segment for receiving a separated telephonic communication, the separated telephonic communication comprising at least one of first constituent voice data and second constituent voice data;
a code segment for receiving an input transmission of a gender detection model, the code segment for generating a gender detection model comprising:
a code segment for receiving predetermined male and female voice data; and,
a code segment for identifying aural characteristics and frequencies associated with the received male and female voice data;
a code segment for identifying at least one aural segment in at least one of the first and second constituent voice data,
a code segment for analyzing at least a portion of one of the first and second constituent voice data by applying a gender detection model to the identified aural segment of the first and second constituent voice data;
a code segment for generating gender detection data corresponding to the analyzed portion of one of the first and second constituent voice data, the gender detection data comprising at least an indicator of the gender of the communicant of the analyzed portion of one of the first and second constituent voice data;
a code segment for extracting frame level data from the identified aural segment;
a code segment for generating an aural data packet having at least one frame, wherein the aural data packet is of a length such that psycholinguistics of the aural data packet is capable of being identified; and,
a code segment for analyzing the aural data packet to generate at least one aural data packet statistic.
2. The non-transitory computer readable medium of claim 1 further comprising a code segment for separating the telephonic communication into at least the first constituent voice data and the second constituent voice data.
3. The non-transitory computer readable medium of claim 2 , wherein the communication has a communication protocol, and wherein the code segment for separating the telephonic communication into at least the first constituent voice data and the second constituent voice data further comprises:
a code segment for recording the first constituent audio data on a first audio track of a data file based on the communication protocol; and,
a code segment for recording the second constituent audio data on a second audio track of the data file based on the communication protocol.
4. The non-transitory computer readable medium of claim 1 , wherein the code segment for extracting the frame level data further comprises:
a code segment for decomposing the identified aural segment into at least one frame having a stationary aural signal wherein the decomposing generates at least one frequency;
a code segment for attenuating at least one generated frequency;
a code segment converting the attenuated generated frequency to frequency domain; and,
a code segment for applying a non-linear filter to the attenuated generated frequency in the frequency domain such that the non-linear filter emphasizes at least one aural frequency.
5. The non-transitory computer readable medium of claim 1 , wherein the aural data packet statistic is configured to identify the gender classification of the aural data packet.
6. The non-transitory computer readable medium of claim 5 further comprising:
a code segment for analyzing a plurality of the aural data packets, wherein each aural data packet comprises an aural data packet statistic identifying the gender classification of the respective aural data packet;
a code segment for counting the number of aural data packets corresponding to a predetermined gender classification;
a code segment for comparing the number of aural data packets corresponding to a predetermined gender classification to a predetermined threshold; and,
a code segment for identifying the gender of the communicant based on the comparison.
7. The non-transitory computer readable medium of claim 1 further comprises a code segment for generating predictive customer behavioral data based on an analysis of at least the gender detection data.
8. The non-transitory computer readable medium of claim 1 further comprising:
a code segment for identifying a plurality of the aural segments wherein each aural segment corresponds to at least one word spoken by a communicant;
a code segment for decomposing the plurality of identified aural segments into a plurality of respective frames;
a code segment for generating a respective frequency for each of the plurality of identified and decomposed aural segments;
a code segment for converting each respective generated frequency for each aural segment to frequency domain; and,
a code segment for applying a weighted neural model to the plurality of frames wherein the gender detection data is generated based on the applying of the weighted neural model.
9. The non-transitory computer readable medium of claim 1 further comprising a code segment for generating predictive behavioral data corresponding to the communicant based on the applying of the gender detection model.
10. A non-transitory computer readable medium adapted to control a computer and comprising a plurality of code segments for determining the gender of a communicant in a communication, the non-transitory computer readable medium comprising:
a code segment for receiving a separated telephonic communication, the separated telephonic communication comprising at least one of first constituent voice data and second constituent voice data;
a code segment for analyzing at least a portion of one of the first and second constituent voice data by applying a gender detection model to at least a portion of one of the first and second constituent voice data;
a code segment for generating gender detection data corresponding to the analyzed portion of one of the first and second constituent voice data, the gender detection data comprising at least an indicator of the gender of the communicant of the analyzed portion of one of the first and second constituent voice data;
a code segment for generating an aural data packet having at least one frame, wherein the aural data packet is of a length such that psycholinguistics of the aural data packet is capable of being identified; and,
a code segment for analyzing the aural data packet to generate at least one aural data packet statistic.