| Technology project evaluation is a complex task, it needs massive and specialized knowledge. This task presents a great challenge to assessors and personnel. So it is possible to build an assistant decision-making system for technology project review based on the Internet search technology, intelligent information processing and natural language processing technology.Based on the above objects, this paper aims to design and implement an assistant decision-making system for technology project review. It focuses on optimizing the algorithm in each module of system. The specific contents are as follows:1.This paper focus on exploring filter policies to ensure that web crawler is only to crawl technology pages and ignore other topic pages. The system also needs to complete the duplicate removal work of URL and text. This paper refines a method of near-duplicate detection for text content based on semantic similarity. It can realize fast and efficient access to content on the Internet without repeating.2.For acquired web documents, the semantic analysis technique can transform unstructured or semi-structured data into structured data. With this method, it can extract useful information and strip out unwanted irrelevant information.3.For useful web content data, it first was processed one the language level. Then it organized with an inverted index structure of supporting fast queries.4.This paper also studies influence factors and their weights allocation method to make a mutual ranking. The mutual ranking is about technology project application units and research areas. The application units may include enterprises, universities and research institutes.According to the above research findings, this paper develops an assistant decisionmaking system for technology project review. With the preliminary application, the systems can greatly reduce the workload of assessors, and effectively enhance the decision-making and management level. |
PDF Full Text Request