阅读说明：本技术 通过实时音频流分析解决司乘安全及提升服务质量的方法 (Method for solving driver and passenger safety and improving service quality through real-time audio stream analysis ) 是由 董建鹏 于 2021-07-01 设计创作，主要内容包括：本发明公开了通过实时音频流分析解决司乘安全及提升服务质量的方法,包括以下步骤：订单服务开始时,进行环境音频采集,得到并上传音频流信息；订单服务过程中,收到上传的音频流信息,解析音频流信息,并到敏感词库进行匹配,得到并上传匹配结果,匹配结果上传到安全预警处理系统,触发相对应的安全行为,本发明适用于网约车技术领域,通过实时音频流分析解决司乘安全及提升服务质量,可以实时自动触发安全处理行为,大大降低潜在安全风险,同时能够节省平台处理司乘纠纷的人力和时间成本,实时监督并提醒司乘双方,提高司乘双方的自我约束力,还能提高纠纷处理的效率和公平性,实用性强。(The invention discloses a method for solving driver and passenger safety and improving service quality through real-time audio stream analysis, which comprises the following steps: when order service starts, acquiring environmental audio, and acquiring and uploading audio stream information; the invention is suitable for the technical field of network appointment, solves the driver and passenger safety and improves the service quality through real-time audio stream analysis, can automatically trigger the safety processing behavior in real time, greatly reduces the potential safety risk, can save the manpower and time cost for processing driver and passenger disputes by a platform, supervises and reminds both the driver and passenger in real time, improves the self-restraint of both the driver and passenger, and can improve the efficiency and fairness of dispute processing, and has strong practicability.)

1. The method for solving the driver and passenger safety and improving the service quality through the real-time audio stream analysis is characterized by comprising the following steps of:

when order service starts, acquiring environmental audio, and acquiring and uploading audio stream information;

in the order service process, receiving uploaded audio stream information, analyzing the audio stream information, matching the audio stream information in a sensitive word bank, and obtaining and uploading a matching result, wherein the sensitive word bank comprises a plurality of sensitive words with different grades;

and uploading the matching result to a safety early warning processing system, and triggering corresponding safety behaviors, wherein the safety early warning processing system comprises a plurality of safety behaviors corresponding to a plurality of different grades of sensitive words respectively.

2. The method for resolving ride safety and improving quality of service through real-time audio stream analysis as recited in claim 1, further comprising the steps of:

after the order is finished, receiving driver evaluation and passenger evaluation, wherein the evaluation comprises good evaluation and bad evaluation;

and if the poor assessment occurs, automatically analyzing the audio stream information, assessing the poor assessment information, and recording the poor assessment information for reference as a judgment standard when the driver and the passenger have disputes or complaints.

3. The method for solving driver and passenger safety and improving service quality through real-time audio stream analysis according to claim 1, wherein the audio acquisition of environment is performed at the beginning of the order service to obtain and upload audio stream information, comprising:

when order service starts, a microphone is automatically started to collect environmental audio, audio stream is uploaded to a platform server, the audio sampling rate is 1600Hz or 8000Hz, the sampling precision is 16bits, a sound channel is single channel, an uploading data packet is 1:1 real-time rate, the 8k sampling rate is 640 bytes, the 16k sampling rate is 1280 bytes, and the uploading data is encrypted through an encryption signature algorithm preset by the platform.

4. The method for solving driver and passenger safety and improving service quality through real-time audio stream analysis as claimed in claim 3, wherein in the order service process, the uploaded audio stream information is received, the audio stream information is analyzed, and the matching is performed in the sensitive lexicon to obtain a matching result, comprising:

after the audio stream is uploaded to the platform server, the platform server analyzes the audio file according to the audio stream slice, calls a sensitive word bank of the platform, matches the analyzed content, tags and marks the audio stream according to the safety risk level and the safety event tag of the sensitive word bank if the sensitive word is matched, and then transmits the safety tags and the risk id to the safety early warning processing system.

5. The method for resolving ride safety and improving quality of service through real-time audio stream analysis as claimed in claim 4, wherein the matching result is transferred to a safety precaution processing system to trigger corresponding safety actions, comprising:

the safety early warning processing system automatically distributes corresponding reminding and warning contents based on the safety label and the risk id, or automatically triggers to dial an alarm call, or automatically dials an emergency contact person, or automatically establishes contact between the platform and a driver or a passenger for manual intervention.

6. The method for resolving ride safety and improving quality of service through real-time audio stream analysis as recited in claim 5, wherein: the safety early warning system can collect the data of each order, deliver the data to the deep learning system, continuously analyze and learn the order information of the safety event, and update the sensitive word bank.

Technical Field

The invention belongs to the technical field of network appointment, and particularly relates to a method for solving driving and riding safety and improving service quality through real-time audio stream analysis.

Background

The online taxi appointment, namely the short name of the online taxi appointment operation service, refers to the operation activities of establishing a service platform by relying on the internet technology, accessing vehicles and drivers meeting the conditions, and providing non-cruise taxi appointment service by integrating supply and demand information;

because the safety problems of the network car booking related service industry occur frequently in recent two years, the safety and service quality problems in the process of the driver and passenger service are very important to solve; most of the existing technologies collect the recording in the service process, and then the recording file is analyzed after a problem exists; if a safety problem occurs in the service process, if an alarm is required, the user also needs to operate the mobile phone to initiate the alarm; the modes cannot automatically and quickly solve the safety problem in real time, and the conventional network appointment service evaluation mode is basically mutual evaluation of drivers and conductors, and then if disputes exist, the drivers and conductors contact the platform and manually intervene to carry out repeated evaluation, so that the mode is time-consuming and labor-consuming.

Disclosure of Invention

The invention aims to overcome the defects of the prior art and provide a method for solving riding safety and improving service quality through real-time audio stream analysis.

In order to achieve the purpose, the invention adopts the following technical scheme:

the method for solving the driver and passenger safety and improving the service quality through the real-time audio stream analysis comprises the following steps:

when order service starts, acquiring environmental audio, and acquiring and uploading audio stream information;

in the order service process, receiving uploaded audio stream information, analyzing the audio stream information, matching the audio stream information in a sensitive word bank, and obtaining and uploading a matching result, wherein the sensitive word bank comprises a plurality of sensitive words with different grades;

and uploading the matching result to a safety early warning processing system, and triggering corresponding safety behaviors, wherein the safety early warning processing system comprises a plurality of safety behaviors corresponding to a plurality of different grades of sensitive words respectively.

Preferably, the method further comprises the following steps:

after the order is finished, receiving driver evaluation and passenger evaluation, wherein the evaluation comprises good evaluation and bad evaluation;

and if the poor assessment occurs, automatically analyzing the audio stream information, assessing the poor assessment information, and recording the poor assessment information for reference as a judgment standard when the driver and the passenger have disputes or complaints.

Preferably, when the order service is started, the acquiring of the environmental audio is performed, and the obtaining and uploading of the audio stream information includes:

when order service starts, a microphone is automatically started to collect environmental audio, an audio stream is uploaded to a platform server through a websocket protocol, the supported audio formats include pcm, wav, mp3 and aac, the request protocol adopts a wss protocol, the audio sampling rate is 1600Hz or 8000Hz, the sampling precision is 16bits, a sound channel is a single channel, an uploading data packet is at a 1:1 real-time rate, the 8k sampling rate is 640 bytes, the 16k sampling rate is 1280 bytes, and the uploading data is encrypted through an encryption signature algorithm preset by the platform.

Preferably, in the order service process, the receiving of the uploaded audio stream information, the parsing of the audio stream information, and the matching to the sensitive lexicon to obtain the matching result includes:

after the audio stream is uploaded to the platform server, the platform server analyzes the audio file according to the audio stream slice, calls a sensitive word bank of the platform, matches the analyzed content, tags and marks the audio stream according to the safety risk level and the safety event tag of the sensitive word bank if the sensitive word is matched, and then transmits the safety tags and the risk id to the safety early warning processing system.

Preferably, the matching result is transferred to a safety early warning processing system to trigger a corresponding safety behavior, including:

the safety early warning processing system automatically distributes corresponding reminding and warning contents based on the safety label and the risk id, or automatically triggers to dial an alarm call, or automatically dials an emergency contact person, or automatically establishes contact between the platform and a driver or a passenger for manual intervention.

Preferably, the safety early warning system collects data of each order, delivers the data to the deep learning system, and continuously analyzes and learns the order information of the safety event to update the sensitive word bank.

In summary, due to the adoption of the technical scheme, the invention has the beneficial effects that:

the invention solves the driver and passenger safety and improves the service quality through real-time audio stream analysis, can automatically trigger the safety processing action in real time, greatly reduces the potential safety risk, can save the manpower and time cost for processing the driver and passenger disputes by a platform, processes most of work through cloud automation, can supervise and remind both drivers and passengers in real time through real-time speech analysis, improves the self-restraint force of both drivers and passengers, creates a harmonious network appointment experience with high-quality service, can automatically process, evaluate and record the driver and passenger evaluation after the order is finished, provides the judgment standard for the subsequent driver and passenger dispute, improves the efficiency and fairness of dispute processing, and has strong practicability.

Drawings

FIG. 1 is a flow chart of the method for solving riding safety and improving service quality through real-time audio stream analysis according to the present invention.

Detailed Description

The following describes a specific embodiment of the method for solving riding safety and improving service quality through real-time audio stream analysis according to the present invention with reference to fig. 1. The method for solving the riding safety and improving the service quality through the real-time audio stream analysis is not limited to the description of the following embodiments.

Example 1:

the embodiment provides a specific structure of a method for solving driver and passenger safety and improving service quality through real-time audio stream analysis, as shown in fig. 1, the method includes the following steps:

when order service starts, acquiring environmental audio, and acquiring and uploading audio stream information;

in the order service process, receiving the uploaded audio stream information, analyzing the audio stream information, matching the audio stream information in a sensitive word bank, and obtaining and uploading a matching result, wherein the sensitive word bank comprises a plurality of sensitive words with different grades;

and uploading the matching result to a safety early warning processing system, and triggering corresponding safety behaviors, wherein the safety early warning processing system comprises a plurality of safety behaviors corresponding to a plurality of different grades of sensitive words respectively.

Specifically, the method further comprises the following steps:

after the order is finished, receiving driver evaluation and passenger evaluation, wherein the evaluation comprises good evaluation and bad evaluation;

and if the poor assessment occurs, automatically analyzing the audio stream information, assessing the poor assessment information, and recording the poor assessment information for reference as a judgment standard when the driver and the passenger have disputes or complaints.

Specifically, when the order service is started, the environmental audio acquisition is performed, and the audio stream information is obtained and uploaded, including:

when order service starts, a microphone is automatically started to collect environmental audio, an audio stream is uploaded to a platform server through a websocket protocol, the supported audio formats include pcm, wav, mp3 and aac, the request protocol adopts a wss protocol, the audio sampling rate is 1600Hz or 8000Hz, the sampling precision is 16bits, a sound channel is a single channel, an uploading data packet is at a 1:1 real-time rate, the 8k sampling rate is 640 bytes, and the 16k sampling rate is 1280 bytes, and the uploading data is encrypted through an encryption signature algorithm preset by the platform, so that the security of the collected data is ensured.

Specifically, in the order service process, the method receives the uploaded audio stream information, analyzes the audio stream information, and matches the audio stream information with a sensitive word bank to obtain a matching result, and includes:

after the audio stream is uploaded to the platform server, the platform server analyzes the audio file according to the audio stream slice, calls a sensitive word bank of the platform, matches the analyzed content, tags and marks the audio stream according to the safety risk level and the safety event tag of the sensitive word bank if the sensitive word is matched, and then transmits the safety tags and the risk id to the safety early warning processing system.

Further, the matching result is transferred to a safety early warning processing system, and a corresponding safety behavior is triggered, including:

the safety early warning processing system automatically distributes corresponding reminding and warning contents based on the safety label and the risk id, or automatically triggers to dial an alarm call, or automatically dials an emergency contact person, or automatically establishes contact between the platform and a driver or a passenger for manual intervention.

Furthermore, the safety early warning system can collect the data of each order, deliver the data to the deep learning system, continuously analyze and learn the order information of the safety event, and update the sensitive word bank.

The working principle is as follows: as shown in fig. 1, firstly, when an order starts to serve, a driver APP starts audio acquisition, driver audio streams are uploaded to a platform server in real time, the platform server stores the order audio streams, meanwhile, an audio stream analysis system starts to work, analyzes audio stream information and matches the audio stream information to a sensitive word bank (the sensitive word bank stores information such as a security event label), a security processing event is automatically triggered by a matching result, security event classification is performed, if the order is normally finished, driver evaluation is performed after the order is finished, a scoring and checking system is triggered to work, if the order is a general security event, a driver is reminded to pay attention to words through voice or pop windows in the driver APP, and if the order is an emergency security event, an alarm or a contact telephone is automatically dialed.

The safety early warning system can collect the data of each list, the data are sent to the deep learning system, order information of safety events is continuously analyzed and learned, and a set of more detailed and complete sensitive word bank is formed, so that a safety problem grading system is more accurate, and a safety problem automatic processing system is more refined. Through continuous learning, the problem processing capability of the system is improved, and more labor and event costs are reduced.

At present, the traditional network car booking service evaluation mode is basically that drivers and conductors mutually evaluate, and then if disputes exist, the drivers and conductors contact the platform and manually intervene to conduct reply judgment on the evaluation. This mode is time consuming and labor intensive. And the platform scoring and auditing system can trigger the auditing operation immediately after the order evaluation is submitted. And aiming at poor evaluation and the like, automatically extracting the order safety problem labels recorded in the safety early warning system, and analyzing the authenticity and reliability of evaluation. If the order is matched with the safety problem label, automatically extracting the safety problem label and the related audio file in the safety early warning system, and establishing a grading and inspection file of the order in an evaluation system. And displaying the key information to the scored person. If the department has a dispute, the system can extract the scoring and disputing archive as key evidence to judge the responsibility.

The invention solves the driver and passenger safety and improves the service quality through real-time audio stream analysis, can automatically trigger the safety processing action in real time, greatly reduces the potential safety risk, can save the manpower and time cost for processing the driver and passenger disputes by a platform, processes most of work through cloud automation, can supervise and remind both drivers and passengers in real time through real-time speech analysis, improves the self-restraint force of both drivers and passengers, creates a harmonious network appointment experience with high-quality service, can automatically process, evaluate and record the driver and passenger evaluation after the order is finished, provides the judgment standard for the subsequent driver and passenger dispute, improves the efficiency and fairness of dispute processing, and has strong practicability.

The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.