Printer Friendly

Design of the information service push system for prevention and control of fruit trees diseases and insect pests.

1. Introduction

China has a long history in the cultivation of fruit trees and the planting structure is diverse. Fruit tree planting is the third largest industry in planting industry, which is of great significance in improving peasants' income and developing modern agriculture. Diseases and insect pests severely compromise fruit tree yield and fruit quality. Every year famers invest large amount of energy and money into the prevention and treatment of diseases and insect pests. With the development and popularization of computer network, farmers started to obtain prevention and control information of diseases and insect pests through the Internet (Leite et al., 2016). However, the information obtained has a lot of varieties and poor pertinence various, which cannot effective solve the diseases and insect pests issue in orchards. Therefore, it is an urgent need to establish the information push system of fruit trees diseases and insect pests to collect and analyze the new technology and new reagent of prevention and control of fruit trees diseases and insect pests. The interest and behavior model of famers needs to be established to actively push the optimal prevention and control scheme of diseases and insect pests, realizing personalized information push service of fruit trees diseases and insect pests and promoting the information sharing.

In recent years, the research on personalized agricultural information push service has made some progress in our country. RSS agricultural information push service (Li et al., 2006) and the user interest model of RSS personalized information service (Guo et al, 2010) can be adopted to achieve convenient and quick push of latest information. The ontology-based personalized push model of agricultural information service (Qiao et al, 2013) and the personalized recommendation algorithm based on ontology user interest model (Yan et al., 2010; Prieto et al., 2014) can improves the accuracy of personalized recommendation. Aiming at the problems of low sharing rate of agricultural scientific and technological information and dispersed distribution, modified K-means clustering method to establish user access model and studied the personalized service push model of agricultural scientific and technological information (Xiao et al, 2013), the targeted agricultural information service according to the factors like region and time and created personalized multimedia agricultural information push system (Chen et al., 2014). The above mentioned researches promote the spread and development of agricultural scientific and technological information in our country. 2

2. System Framework

This system adopted B/S structure, realizing the automatic acquisition of prevention and control information of fruit trees diseases and insect pests and personalized information push. On the client side, farmers can access the information resources in this system through browser and they can choose different push methods like SMS, RSS and e-mail. On the server side, focused crawler technology is utilized to automatically acquire the information diseases and insect pests and store it into local database; data mining technology is adopted to analyze user logs and Cookies. Combined with the user registration information, the interest orientation of users is determined to establish user interest model and push the prevention and control information to farmers based the model. This system is classified into data layer, logic business layer and presentation layer and the system functional structure is shown in Figure 1.

3. Key Technology of Personalized Push

3.1. User Interest Model

In general, user interest can be classified into short-term and long-term interest. Shortterm interest is stimulated by environmental conditions and the long-term interest is produced by subjective tendency. The acquisition modes of user interest can be classified into explicit and implicit acquisition. This research adopts the way of combining explicit and implicit acquisition to build user interest model. Explicit building of interest model can be obtained mainly through user registration information and the inexplicit upgrading interest model can be obtained mainly through the mining of the content and behavior in browsing the web of users (like logs, Cookie). Through the analysis of user preference, combined with the feedback of users to the push information, conduct upgrading of user interest model, the flowchart of user interest model building is shown in Figure 2.

3.1.1. Building of User Interest Model

The user interest model consists of the objects representing user interest and the interest level of users in this information can be expressed by weighted value. The user interest model in this research is represented by key words and the vector space model of weight. TF-IDF algorithm is adopted to calculate the weight of key words in the document. The formula is as follow:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

TF ([k.sub.i], d) is the number of occurrences of key word [k.sub.i] in document d ; DF ([k.sub.i]) is the number of documents containing the key word; n is the total number of documents. After the calculation of the weight of key words, the topic vector of user interest can be obtained.

Use a group of interest key words [I.sub.1], [I.sub.2], ... [I.sub.n] to represent n interest models of users. The weighting [w.sub.i] of each interest model [I.sub.i] can be conducted based on user interest level. The user interest can be expressed as the weighted vector model of (([I.sub.1], [w.sub.1]), ([I.sub.2], [w.sub.2]), ... ([I.sub.n], [W.sub.n])).

3.1.2. Upgrading of User Interest Model

The upgrading of user interest model mainly considers from short-term upgrading and long-term upgrading (YI Wen-wen et. al, 2010). The upgrading strategy of short-term upgrading is that: [I.sub.1], [I.sub.2], ... [I.sub.n] are all leaf node sets in the user interest model and P represents the page that users access. When satisfying the following conditions, P belongs to interest type [I.sub.i].

S (P, [I.sub.i]) = 1/[absolute value of [I.sub.i]] [summation over ([I.sub.i] [member of] I)] S(P, [I.sub.j]) (2)

S(P, [I.sub.i]) > S(P, [I.sub.j])j = 1,2, ..., n, j [not equal to] i 9 (3)

S(P, [I.sub.i]) represents the similarity of page P and interest [I.sub.i]. When the similarity of page P and interest [I.sub.i] reaches its maximum, add the vector of P into the [I.sub.i] and upgrade the information of all nodes from the root node to the leaf node.

3.1.3. Introduction of Forgetting Mechanism

When upgrading the user interest model, we not only need to add the newest interest feature words of the users, but adjust the existing weight of feature words in the model. We need to introduce forgetting factor [K.sub.x] to modify, which is to multiply the weight of feature words with [K.sub.x] to improve the accuracy of recommendation. The forgetting [K.sub.x] is:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)

Tnow is the current date; Tcreat represents the date of first appearance of the interest feature word in the model; Tupdate is the update time of interest feature word; Tnow - Tupdate is the half-life period (half period of user interest).

3.2. Information Push Technology

Information push technology is a kind of information acquisition technology, which is a new-type information transmission system that the information service company or network company can obtain the information from the information resource or the information processor through certain technical standard or protocol and then send SMS through fixed channel. Information push technology possesses the characteristic of initiative, individuation, convenience, effectiveness and intelligence. The working flow chart is shown in Figure 3.

3.2.1. Prevention and Control of Diseases and Insect Pests Push According to Fruit Tree Species

China has a great variety of fruit trees, mainly including banana, apple, orange, pear, grape, pineapple, jujube and persimmon. In 2010, the planting area of fruit trees covered 11,543,900 hectares and the yield reached 128,652,000 tons. Every year, the yield reduction due to diseases and insect pests is 25% and the direct economic loss reaches up to billions of Yuan (Du, 2005). The prevention and control of diseases and insect pests is dominated by chemical pesticides. Farmers use pesticides blindly, lacking scientific guidance. Aiming at the diseases and insect pests of different fruit tree species, based on the record of browsing system page of famers, the potential interest information of users is mined and user interest models are built to push the prevention and control information of diseases and insect pests effectively and precisely, improving the personalized level of push service.

3.2.2. Pushing Prevention and Control Scheme According to the Critical Period of Fruit Tree Management

Fruit tree planting has obvious seasonal and periodic characteristics and the occurrence of diseases and insect pests is closely related to the growth period of fruit trees. In different seasons, the objects, schemes, reagents, and key technology of diseases and insect pests prevention and control are different. According to the occurrences of diseases and insect pests in the key management period of fruit trees, combined with the accessing time of farmers, farmers' prevention and control information of interest of diseases and insect pests is determined and corresponding prevention and control schemes are pushed, improving the pertinence and effectiveness of information push of the system.

4. Design and Realization of the System

4.1. System Function Design

4.1.1. User Access Analysis Module

This module is mainly used for the analysis of user access information. Through the analysis and reorganizing of access logs on the client-side, combined with page key words, Html document information is extracted and user interest model is built applying TF-IDF method to calculate the weight. Upgrading of user interest model is conducted according to the method in 2.1 to provide basis for user classification.

User classification model can be defined as User = {F,I,DI,T,CT} . F represents fruit tree species, like banana, apple, pear; I represents the set of user interest information and I = {[I.sub.1], [I.sub.2], ... [I.sub.n]}; DI represents users' interest level of the information; T is users' attention level of the information and 0,1,2,3 is applied to represent no attention, little attention, attention and much attention; CT represents the documents and images of interest of users.

4.1.2. Information Acquisition Module

This module is mainly used for the acquisition of fruit tree prevention and control information of diseases and insect pests. Crawler technology is applied to acquire the pages related to the occurrence, prediction and forecast, prevention and control, reagent of fruit tree diseases and insect pests from relevant information source websites, then download them and save on the local server. Web information extraction technology is adopted and html document is parsed into DOM tree. And then, the delimiter of the target data item is determined based on nodes of DOM tree. The data item is determined and the information needed is extracted, which is later transformed into XML document of form. In addition, remove irrelevant html tag and CSS style. Finally, transform it into the XML document and store it into the information database of diseases and insect pests.

4.1.3. Information Push Module

This module is mainly used for the processing of user access requests. According to the interest information inputted by the user, obtain implicit feedback information by tracing the access operation of and mining the visited pages of the user. Based on user interest need, formulate relevant prevention and control schemes of diseases and insect pests and push the information to terminals including computer, smartphone through push approaches like e-mail, SMS and RSS to provide sound information service to users.

E-mail push is to send the prevention and control information source of diseases and insect pests to the computer and cellphone of farmers by the e-mail push on a regular basis. SMS push is to send the customized prevention and control information of diseases and insect pests to the cellphone of farmers by the SMS push on a regular basis. RSS channel push is to provide famers with the RSS channel satisfying their interest and the information is sent in the form of metadata abstract. These three push approaches can accept the feedback information of users and then adjustment of push content can be conducted.

4.2.Implementation Process of the System

This system adopts B/S structure to facilitate maintenance and upgrading. Visual studio 2005 is adopted to build the development environment based on .net; SQL Server 2005 is adopted as the background database; and ASP.NET is taken as development language. The system obtains the diseases and insect pests information receiving much attention of farmers through information acquisition module and then stores it into the information database of diseases and insect pests. The system can obtain the latest access record of famers when they visit this system. The interest and hobby of famers are recognized and predicted intelligently and contrast is conducted according to famers' requirement and the related information in the information database of diseases and insect pests. The customized information is searched automatically, and the prevention and control information of diseases and insect pests is pushed to farmer through information push module on a regular basis, thus satisfying personalized needs.

5. Discussion and Conclusion

In recent years, China has gained a series of scientific achievements in the field of fruit tree planting protection. However, some achievements have not been popularized and applied widely, the reason of which is the lacking of effective popularization mechanism and information push method, so the achievements cannot be pushed to the base. This research aims at the issue of prevention and control information push of diseases and insect pests and builds the user interest model adopting the combination of explicit and implicit approach through the mining of web logs and Cookie. The prevention and control schemes of diseases and insect pests are pushed according to fruit tree species and critical period of management. Combined with personalized push technology like e-mail, SMS and RSS, the information service push system for prevention and control of fruit tree diseases and insect pests suitable for the interest needs of farmers is built, improving the quality of information service and promoting the development of modern agriculture.

Recebido/Submission: 03/04/2016

Aceitacao/Acceptance: 16/07/2016

Acknowledgements

The research was supported by the Scientific research and development project of Baoding (16ZN006); and by the project of Education Department of Hebei Province (Z2014079)

References

Chen C., Wu H.R. (2014). Personalized Multimedia Agriculture Information Push System. Jiangsu Agricultural Sciences, 42(9): 395-399.

Du Y.B. (2005). Biological Control of Pests of Fruit Trees. Journal of Sichuan Forestry Science and Technology, 26(6): 73-77.

Guo L.J. (2010). User Interest Model of Rss Personalized Information Service Research and Application. Beijing University of Chemical Technology.

Leite, P., Goncalves, J., Teixeira, P., & Rocha, A. (2016). A model for the evaluation of data quality in health unit websites. Health Informatics Journal, 22(3): 479-495.

Li Q.W., Wang W.S., Zhou Z.Y. (2006). The research and implement of agriculture information push technology based on RSS. Agriculture Network Information, (4): 10-12.

Prieto, A.E., Lozano-Tello, A. (2014). Adaptacion de Workflows basada en Ontologias. RISTI--Revista Iberica de Sistemas e Tecnologias de Informacao, (14): 67-82.

Qiao bo, Nie Xiao-Yi, Fang Kui. (2013). Ontology-based Agricultural Information Service Personalized Push Research. Journal of Anhui Agri. Sci.41, (27): 11213-11214.

Xiao H., Li S.Q., Wang C. (2013). Research on Personalized Service Push Model for Agricultural Science and Technology Information. Computer Applications and Software, 30(6): 93-96, 136.

Yan J.W., Huang X., Liu M., Zhu Y.B., Ni H.B. (2010). Personalized recommendation algorithm for user interest model based on ontology. Computer Integrated Manufacturing Systems, 16(12): 2757-2762.

Yi W.W., He F.N. (2010). Representation and Update for User Interest Model Based on User Browsing Behaviors. Journal of Changzhou Vocational College of Information Technology, 9(4): 28-30.

Qu Yun (1), Tao Bu (2)]

5469831@qq.com, 331303838@qq.com

(1) Academic Affairs Office, Agriculture University of Hebei, 071001, Baoding, China

(2) College of Plant Protection, Agriculture University of Hebei, 071001, Baoding, China
COPYRIGHT 2016 AISTI (Iberian Association for Information Systems and Technologies)
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Yun, Qu; Bu, Tao
Publication:RISTI (Revista Iberica de Sistemas e Tecnologias de Informacao)
Date:Oct 15, 2016
Words:2690
Previous Article:Study on the effect of customer participation on innovation performance in Chinese IT industry.
Next Article:A stochastic dominance test for the efficiency of program trading algorithm.
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters