The rapid expansion of information content on the internet and the mobile internet has resulted in violations of laws and regulations and bad information,which affects the content security of the internet space.Traditional text content security recognition methods based on matching of sensitive words ignore context semantics,resulting in high false positive rate and low accuracy.Based on the analysis of traditional text content security recognition methods,a fusion recognition model using deep learning and a model fusion algorithm process were proposed.Text content security recognition system based on the fusion recognition model using deep learning and experimental verification was introducted deeply.Results show that the proposed model can effectively solve the problem of high false positive rate caused by the lack of semantic understanding of traditional recognition methods,and improve the accuracy of the bad information detection.

With the obvious improvement of various telecom business demands,the outbound call industry,which is the main channel for marketing,resolution and notification of telephone services under the traditional mode,is facing more and more pain points,including market,personnel,training,customer sentiment,cost,accurate data and people flow in all aspects.In order to realize the cost reduction and efficiency increase,inject wisdom and improve capability of telecom services,the main needs of the existing network outbound call system were combined with the application of artificial intelligence and big data technology in the field of outbound calls,aim to build intelligent and high efficiency.A discussion of the feasibility of the program was given based on the network construction involved in the efficiency of the ＆quot;smart outbound system＆quot;.

Text classification is very important to text data mining and value exploration.The traditional text classification system has problems of weak feature extraction ability and low classification accuracy.Compared with the traditional text classification technology,deep learning technology has many advantages such as high accuracy and effective feature extraction.Therefore,it is necessary to apply deep learning technology to the text classification system to solve the problems of the traditional text classification system.The traditional text classification system was analyzed,and the architecture and key technologies of text classification system based on deep learning were proposed.Finally,several classification models were verified and compared,including the traditional classification model,TextCNN and CNN+LSTM.

Networking and intelligentization are the future development directions of interactive service systems.Internet popularization and rich social channels have realistic needs for intelligent interactive services.High-quality and low-cost intelligent interactive solutions have become the main issues that need to be addressed in various industries.Relying on artificial intelligence technology,intelligent means such as intent recognition,semantic understanding,dialogue and judgment were introduced,the integration of scene design and knowledge fragmentation processing technology has brought new means and opportunities for the realization of interactive service intelligence.The problems faced by the existing interactive system were analyzed.On this basis,the key functional elements and system architecture of the intelligent interactive system were designed.

As the most important contact point for customers in the internet industry,customer service orders have become the main issues that need to be addressed and urgently solved in various industries.The intelligent sequencing of work orders provides a feasible application for solving this problem.The rapid development of artificial intelligence technology and big data technology has given new directions and opportunities to solve such problems.The background of the intelligent sequencing of customer service orders was analyzed firstly.On the basis of this,the main methods and implementations of intelligent sequencing of customer service orderswere analyzed,and the application of intelligent sequencing of customer service orders based on multi-factor vectorization was expounded.

The big data resources possessed by telecom operators are usually distributed in many different systems,such as DPI、OIDD、CRM.Moreover,the formulation,interpretation and rules of the big data are not always the same in different systems.Therefore,it is difficult to identify and utilize the same object’s multi-type data in different sys-tems.Big data analysis’ sample size and dimension are limited,with the decreasing of analysis results’ reality and accuracy.The methods,architectures and implementation examples of big data’s heterogeneous association were pre-sented.The data fusion in user-dimension from different systems could optimize the data sample space of applications,such as user portrait.Thus,the value of carrier’s big data density was greatly improved.

Device-to-Device（D2D）communication is an important technology in 5G network.It helps to improve the transmission efficiency of data dissemination services.However,due to the extra signaling overhead caused by D2D communications,currently most of enhanced-multicast solutions cannot solve the contradiction between high-efficient data transmission and low-efficient signaling procedure,so as to reduce the overall performance gain.An intra-cluster D2D information sharing algorithm based on matrix analysis which includes centralized and distributed modes was proposed.The proposed algorithm could significantly reduce D2D retransmission times in clusters by adaptively selecting the optimal transmitters and data packets for D2D multicast retransmission,achieving the aim of reducing signaling overhead and transmission delay.

A cloud call center architecture was presented and the call center platform cloud technology and business model were studied. The cloud call center application prospect and business trends were also described. A reference of call center platform cloud and technology development was provided for the operator.

Open-source VoIP technologies can be employed to setup VoIP platforms by deploying open-source software on x86 servers, which have many advantages such as low cost, flexibility and openness. The usage of open-source VoIP technologies in NGN service platform deployment helps to reduce carrier's investment, shorten service deployment cycle and improve the duplicability of service platform. In the paper, the architecture of NGN service platform deployed by open-source VoIP technologies is proposed based on comprehensive survey and detailed analysis. A typical example is also provided to evaluate the feasibility and advantages of the proposed NGN service platform deployment method.

In order to control the QoS situation of voice services in NGN,it is necessary to monitor the voice services in NGN.However,till now,no exact defination for monitoring indexes of NGN voice-service QoS exists in the industry.This paper,based on the demand in monitoring voice-service QoS in NGN,and from three aspects incluling performancd of session control,the quality of voice,the statistics of audio code,defines a set of indexes for monitoring Voice-Service QoS in NGN,also called monitoring parameters for Voice-Services QoS in NGN.The monitoring of these parameters would timely reflect,the Qos of voice service and the performance of network,and provide useful information and data for the ensurance of voice-service QoS guarantee.

The multi-core processor and the multi-core processor based high-performance systems are experiencing a rapid progress due to the fact that the parallel operation ability of processor becomes more and more important in the current IT field.This paper studies the software framework for concurrent computing applications on heterogeneous multi-core system.The proposed framework can accelerate the concurrent applications and makes the application programming on the heterogeneous multi-core system more simply.The Voice Conference prototype using such framework on Cell processor based system for Telecomm application shows good performance which justifies the efficiency of the framework and the potential of the Cell processor for Telecomm application.

在目前运营商运营的呼叫中心系统中,呼叫转出计费大多采用由大网生成话单的方式来实现。随着呼叫中心业务的日益复杂,该方式的不合理性愈发明显,由呼叫中心自己生成转出话单的需求也越来越迫切。本文在分析了呼叫中心生成呼叫转出话单需求的基础上,以中国电信NGCC架构为例,阐述了呼叫中心实现该话单计费功能的困难,提出了三种基于NGN的呼叫中心呼叫转出计费的解决方案,并对三种方案做了横向比较和实现建议。

本文提出了两种通过平台互联来提供区域性号码百事通业务的新方案,并对实施号码百事通平台互联后的业务合作前景、赢利点及结算模式等进行了分析和说明,为全网性号码百事通业务的发展提供了新思路。