Design And Implementation Of Technical Forum Based On Personalized Search Recommendation

With the rapid development of the Internet and information technology in the 21 st century,technology forums have become a key platform for people to obtain computer expertise.People publish their original articles through technical forums,obtain the content in their favorite professional fields,and interact with industry experts on professional issues to improve their professional skills.The technical forum mainly publishes a large number of technical articles or questions,then users state their opinions through replies.In the past,most of the forums were comprehensive forums,however,the advantages of their breadth could not achieve excellence and perfection in every topic.They provided a large amount of content but failed to search and recommend them effectively,so that users received the content information provided by the forum website passively.Moreover,users’ needs for diversified and personalized content cannot be satisfied,restraining the development of cultural values.In order to meet the market-oriented needs of users and break the limitations of traditional forums,this paper aims to develop a professional technical forum platform based on personalized search recommendations for an Internet company.This forum uses the B/S system architecture in the design aspect,and adopts the form of separation of front and back ends.Specifically,the front desk uses the React.js framework,rendering the page to reduce performance consumption and the long-time page loading time;and the background uses Springboot technology framework to reduce the development time,thereby improving development efficiency.After the forum’s business functions are online for a period of time,in order to solve personalized search recommendations,we use the mainstraim recommendation algorithms,i.e.personalized recall ALS algorithm,the logistic regression LR algorithm,the bm25 algorithm in the mainstream search engine Elasticsearch,and the spark data framework to process the data.We also assign different weights to calculate scores according to the user’s behavior attributes,article ratings,etc.,and model all users’ click data and user behavior interaction data,then recommend their favorite technical articles for users.As result,we improve user satisfaction,and then increase the click-through rate,so that the value of enterprise technology can be developed.This paper first introduces the research background and development trend of the technical forum based on personalized search recommendation;then by comparing the advantages and disadvantages of the current mainstream recommendation algorithms,chooses the hidden semantic model-based ALS algorithm in the matrix factorization collaborative filtering algorithm;then for the demand analysis,based on theoretical research and technical support,designs search service,personalized recommendation service,front-end business system and operation-end business system in detail,including the overall framework design of the forum,the architecture design of search services and recommendation services,the key business functions,Database design and algorithm flow design;Finally,we show the realization result display and test case result display,carried out on the technical forum platform based on personalized search recommendation.