Connection management service
地域：
WA
WA Seattle
摘要
A technology is described for determining a trust score or trust metric used to determine whether to allow a connection to a high risk destination. In one example of the technology, a connection request to initiate a connection with a destination over a network may be received at a communication service. The communication service may identify a customer account associated with the connection request and obtain a trust score that indicates whether the connection request associated with the customer account may be allowed. The communication service may determine whether the connection request associated with the customer account is allowed to initiate the connection with the destination based on the trust score, and initiate the connection with the destination over the network when the trust score allows the connection request.
說明書
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Voice-communication systems allow users who may be remote from one another to communicate over a communications network using an array of different devices. These voice-communication services may allow users to place telephone calls using an application hosted on a client device or a VOIP (voice over IP) phone, and send voice traffic via a private branch exchange (PBX) over the Internet and other networks, including a public switched telephone network (PSTN).
Premium rate telephone numbers are telephone numbers for telephone calls that incur a higher charge as compared to charges for calls to other telephone numbers. A telephone service provider charges the calling party a high rate and shares the revenue with the owner of the premium rate number. Although billing for premium rate telephone calls is different from standard telephone calls, routing of premium rate telephone calls is typically the same as standard telephone calls. Criminals may also use premium rate numbers to defraud unsuspecting users. As one example, a user may be induced to download dialer software that surreptitiously dials a premium rate number in order to accumulate charges on the user's phone bill without the user's knowledge.
權(quán)利要求
What is claimed is:1. A system comprising:at least one processor; and a memory device including instructions that, when executed by the at least one processor, cause the system to:receive, at a communication service hosted in a service provider environment, a connection request to initiate a connection between an origin endpoint at a first location and a destination endpoint at a second location over a public switched telephone network (PSTN), the first location being remote from the second location; identify a customer account associated with the connection request; obtain, based on the connection request, a trust score that indicates whether the connection request associated with the customer account is allowed, wherein the trust score is determined based in part on a risk score associated with the destination endpoint and a pattern of previous voice connections of the customer account with one or more first destination endpoints categorized as a high risk destination endpoint and one or more second destination endpoints categorized as a high cost destination endpoint; determine that the connection request associated with the customer account is allowed to initiate the connection with the destination endpoint based in part on the trust score; and initiate the connection with the destination endpoint over the PSTN. 2. The system in claim 1 , wherein the connection with the destination endpoint is initiated based in part on a lack of deviation of the connection request with respect to the pattern of previous voice connections. 3. The system in claim 1 , wherein the connection request is received from a communications application hosted on a client device that transmits and receives communication data for the connection with the destination endpoint. 4. The system in claim 1 , wherein the instructions, when executed by the at least one processor, further cause the system to:determine that a number of calls to the destination endpoint during a predetermined amount of time is equal to or greater than a predetermined number; and determine the risk score based in part on the number of calls during the predetermined time period being equal to or greater than the predetermined number. 5. The system in claim 1 , wherein the instructions, when executed by the at least one processor, further cause the system to:determine, at a first time, that the risk score meets or exceeds a threshold value; classify, based in part on the risk score meeting or exceeding the threshold value, the destination endpoint as the high risk destination endpoint; and determine, at a second time that is subsequent to the first time and based in part on the risk score being classified as the high risk destination endpoint, the trust score. 6. The system in claim 1 , wherein the instructions, when executed by the at least one processor, further cause the system to:determine that a first calling rate associated with the destination endpoint is a threshold amount greater than second calling rates associated with one or more different destination endpoints; classify, based in part on the first calling rate being the threshold amount greater than the second calling rates, the destination endpoint as the high cost destination endpoint; and determine the risk score based in part on the destination endpoint being classified as the high cost destination endpoint. 7. The system in claim 1 , wherein the instructions, when executed by the at least one processor, further cause the system to:determine one or more factors associated with the destination endpoint, the one or more factors including first calls to the destination endpoint outside of a predetermined range of hours, the destination endpoint being an unrecognized destination endpoint, and a duration of second calls to the destination endpoint being equal to or greater than a predetermined amount of time; and determine the risk score based in part on the one or more factors. 8. A computer implemented method comprising:receiving a connection request to initiate a connection between an origin endpoint and a destination endpoint that is remote from the origin endpoint over a network; identifying a customer account associated with the connection request; obtaining, based on the connection request, a trust score that indicates whether the connection request associated with the customer account is allowed, wherein the trust score is determined based in part on a risk score associated with connecting to the destination endpoint over the network and a pattern of previous voice connections of the customer account with one or more first destination endpoints categorized as a high risk destination endpoint and one or more second destination endpoints categorized as a high cost destination endpoint; determining whether the connection request associated with the customer account is allowed to initiate the connection with the destination endpoint based in part on the trust score; and initiating the connection with the destination endpoint over the network based in part on a determination that the connection request is allowed to initiate the connection with the destination endpoint based in part on the trust score. 9. The method in claim 8 , further comprising:determining that the risk score of the destination endpoint meets or exceeds a threshold value; and identifying a condition linked to the customer account that the trust score be evaluated prior to allowing the connection with the destination endpoint. 10. The method in claim 8 , wherein obtaining the trust score further comprises sending a trust score request to a scoring agent that generates the trust score using: (i) destination risk information for the destination endpoint, and (ii) customer status information and call history information linked to the customer account. 11. The method in claim 8 , wherein obtaining the trust score further comprises:obtaining destination risk information, customer status information, and call history information linked to the customer account; and applying a rule set to the destination risk information, the customer status information, and the call history information to determine the trust score. 12. The method in claim 8 , wherein obtaining the trust score further comprises:analyzing a customer call history linked to the customer account for the pattern of previous voice connections that meet or exceed a risk threshold; and generating the trust score based in part on matching the connection request to the pattern of previous voice connections. 13. The method in claim 8 , wherein obtaining the trust score further comprises:training a machine learning model to generate trust scores that are classifications using training data that includes customer account features for a plurality of customer accounts and destination features for a plurality of destination endpoints; inputting one or more customer account features for the customer account and one or more destination features for the destination endpoint to the machine learning model; and determining, as an output from the machine learning model, a trust score classification for the customer account. 14. The method in claim 8 , further comprising:receiving a second connection request to initiate a second connection between the origin endpoint and the destination endpoint over the network; determining that the second connection request to initiate the second connection with the destination endpoint is not allowed based in part on the trust score; and providing an indication that the second connection request is denied. 15. The method in claim 8 , further comprising:receiving instructions to override a denial of the connection request; and initiating the connection with the destination endpoint. 16. The method in claim 8 , further comprising:receiving an authorization code with instructions to override a denial of the connection request; validating the authorization code; and initiating the connection with the destination endpoint based in part on the authorization code being validated. 17. The method in claim 8 , wherein the connection request is a call forwarding request that redirects a call from a first destination endpoint to a second destination endpoint. 18. The method in claim 8 , wherein the connection request is sent by a communications application hosted on a client device, and the communications application includes functionality for calling destinations. 19. A non-transitory machine readable storage medium including instructions embodied thereon, wherein the instructions, when executed by at least one processor:receive a connection request to initiate a connection between an origin endpoint and a destination endpoint that is remote from the origin endpoint over a network; identify a customer account associated with the connection request; determine that a first calling rate associated with the destination endpoint is a threshold amount greater than one or more second calling rates associated with one or more different destination endpoints; classify, based in part on the first calling rate being the threshold amount greater than the one or more second calling rates, the destination endpoint as a high cost destination endpoint; determine a risk score associated with the destination endpoint based in part on the destination endpoint being classified as the high cost destination endpoint; obtain, based in part on the connection request and the risk score, a trust score for the customer account that indicates whether the connection with the destination endpoint is allowed for the customer account, wherein the trust score is determined based in part on matching the connection request to a pattern of customer connections to other destination endpoints; determine that the connection request associated with the customer account is allowed to initiate the connection with the destination endpoint based in part on the trust score; and initiate the connection with the destination endpoint over the network. 20. The non-transitory machine readable storage medium in claim 19 , wherein the connection request is received at a communication service hosted in a service provider environment, and the communication service is scaled based in part on demand by launching and terminating computing instances used to host the communication service.
意見反饋
使用該功能遇到問題