With the increase of the scale of business process model management database,traditional model management methods are unable to meet the expectations in terms of efficiency and accuracy,and the technology that can improve the efficiency of business process model management has become an urgent demand.Technology of business process similarity can effectively improve efficiency and accuracy of model analysis in scenarios like model search and consistency judge.Therefore,the research on techno-logy of business process similarity has become a research hotspot in the model analysis field.In recent years,researchers have got many valuable achievements,the technologies of business process similarity have developments in many branches involved in different areas.Although there are comparison of technologies in specific branch,there is a lack of systematic research on technologies of business process model similarity.We analyze the calculations of business process model similarity from these dimensions include text similarity,semantic similarity,structure similarity,behavior similarity and human estimation-based similarity,and summarizes the features of these measurements.We find that the technology of business process model similarity is commonly put into these applications include conformance,standardization,search and reuse,then we analyze the research on these scenarios.At last,the challenges of business process model similarity research are analyzed.

Grover quantum search algorithm is a famous quantum algorithm designed for unstructured search problems.It can be used to solve problems such as graph coloring and shortest path sorting,and can also effectively decipher cryptosystems.Graph coloring problem is one of the most famous NP complete problems.In this paper,the graph coloring problem is transformed into an undirected graph in mathematics,and then it is transformed into a Boolean satisfiability problem by using Boolean expression.The steps and principles of solving Boolean expression in quantum circuit diagram and the transformation process from graph co-loring problem to Boolean satisfiability problem are introduced.Finally,on the IBMQ cloud platform,for the three nodes,2-coloring problem and 4-coloring problem are simulated.Experimental results verify the feasibility of using Grover algorithm to solve the graph coloring problem.In the 2-coloring problem with search space of 8 and the 4-coloring problem with search space of 64,the target items are searched with nearly 82% and 97% success probability respectively.In this paper,Grover algorithm is used to solve the 4-coloring problem,expand the experimental scale of the algorithm in this problem field,and improve the quantum circuit of the existing experiments,so that the qubit cost is lower and the result success rate is higher,which shows the remarkable acceleration effect of Grover algorithm in large-scale search problems.

Multi-agent path finding is a classic search problem in the field of artificial intelligence.Conflict-based search algorithm is one of the best algorithms to solve this problem.This paper discusses the basic research of multi-agent path finding,and classifies the research results based on conflict search algorithms and their variants in recent years.According to the improved ways,the variants are divided into four categories,including segmentation strategy improvement,heuristic algorithm,bounded suboptimal algorithm and typical conflict processing.It also introduces the application of the conflict-based search algorithm to the extended problem of multi-agent path finding.Finally,according to the advantages and disadvantages of the current algorithm,the existing challenges are pointed out.In view of these challenges,the possible research directions in the future are given.

The tensor compiler compiles the tensor algorithm and schedule of the operator into the code of the target hardware.In order to accelerate tensor operation,the special processor in the field of deep learning is designed as a special architecture with special instructions,which supports multi-core parallel,multi-level special memory architecture and tensor calculation.On top of the hardware,there is a tensor instruction set closely related to the characteristics of the hardware.In such a complex architecture,the use of tensor instructions has many constraints and limitations,and there are the following problems and challenges.Firstly,the conditional branches introduced by loop tiling such as computing task division or data chunking increase the difficulty of pattern matching.Secondly,tensor instructions have hardware constraints such as alignment and data layout.To solve the above problems and research challenges,an optimization algorithm of tensor instruction ge-neration based on loop partitioning is proposed.By dividing the loop interval,the algorithm eliminates the conditional branches introduced by task division or data segmentation.The instruction and hardware constraints are solved by filling zeros,replacing equivalent instructions and adding additional calculations.The tensor instruction is generated by pattern matching method.This paper studies and extends the open source deep learning compiler TVM version 0.7,and implements a compiler prototype system supporting tensor instruction ge-neration of DianNao architecture machine learning accelerator.In order to evaluate the effectiveness of the algorithm,the operator performance and development efficiency of element-wise binary tensor operator,in-place unary tensor operator and convolution operator are tested on the DianNao architecture machine learning accelerator hardware platform.Experimental results show that the average speedup of the three types of operators is 125.00%,the maximum speedup is 194.00%,and the maximum development efficiency increases by 7 times.