ﻃﺮاﺣﯽ ﻣﺪل ﺧﺪﻣﺎت ﺷﺨﺼﯽﺳﺎزي ﺷﺪه ﺑﺎ روﯾﮑﺮد ﺳﺎﻣﺎﻧﻪﻫﺎي ﺗﻮﺻﯿﻪﮔﺮ در ﻧﺮم اﻓﺰار ﮐﺘﺎﺑﺨﺎﻧﻪ دﯾﺠﯿﺘﺎل آﺳﺘﺎن ﻗﺪس رﺿﻮي

Designing a Personalized Service Model with an Approach to Recommender System in Astan-e Quds-e Razavi Digital Library Software

ﻫﺪف: ﻫﺪف ﭘﮋوﻫﺶ اراﺋﻪ ﻣﺪﻟﯽ ﮐﺎرﺑﺮدي از ﺳﺎﻣﺎﻧﻪ ﺗﻮﺻﯿﻪﮔﺮ )ﭘﯿﺸﻨﻬﺎددﻫﻨﺪه( در ﺣﻮزه ﺧﺪﻣﺎت ﺷﺨﺼﯽﺳﺎزي ﮐﺘﺎﺑﺨﺎﻧﻪ دﯾﺠﯿﺘﺎل آﺳﺘﺎن ﻗﺪس رﺿﻮي اﺳﺖ، اﯾﻦ ﺳﺎﻣﺎﻧﻪ، ﺑﺎ ﺗﺤﻠﯿﻞ اﻟﮕﻮﻫﺎي رﻓﺘﺎري ﮐﺎرﺑﺮان و اراﺋﻪ ﭘﯿﺸﻨﻬﺎدﻫﺎي دﻗﯿﻖ و ﻣﻮرد ﻧﯿﺎز، آﻧﺎن را از ﻣﺮور دادهﻫﺎي ﻏﯿﺮ ﻣﺮﺗﺒﻂ ﺑﯽﻧﯿﺎز ﻣﯽﺳﺎزد. از ﻣﺰاﯾﺎي اﯾﻦ ﺳﺎﻣﺎﻧﻪ ﻣﯽﺗﻮان ﺷﻨﺎﺳﺎﯾﯽ دﻗﯿﻖ ﻧﯿﺎزﻫﺎي اﻃﻼﻋﺎﺗﯽ، اﻓﺰاﯾﺶ ﮐﯿﻔﯿﺖ ﺟﺴﺘﺠﻮ، ﮐﻤﮏ ﺑﻪ اﻧﺘﺨﺎب ﺻﺤﯿﺢ ﻣﻨﺎﺑﻊ اﻃﻼﻋﺎﺗﯽ و ﭘﯿﺶﺑﯿﻨﯽ ﻧﯿﺎزﻫﺎي آﺗﯽ ﮐﺎرﺑﺮان را ﺑﺮﺷﻤﺮد. روش: ﭘﮋوﻫﺶ ﮐﺎرﺑﺮدي اﺳﺖ و ﺑﺎ اﺳﺘﻔﺎده از ﮐﺎوش ﻗﻮاﻧﯿﻦ اﻧﺠﻤﻨﯽ و اﻟﮕﻮرﯾﺘﻢ FP-Growth اﻧﺠﺎم ﺷﺪه اﺳﺖ. ﺟﺎﻣﻌﻪ ﺗﻌﺪاد 960 ﻧﻔﺮ ﮐﺎرﺑﺮ ﻧﺮماﻓﺰار ﮐﺘﺎﺑﺨﺎﻧﻪ دﯾﺠﯿﺘﺎل آﺳﺘﺎن ﻗﺪس رﺿﻮي در ﻓﺎﺻﻠﻪ دو ﺳﺎل ﺑﻮده، ﮐﻪ ﺑﺮ ﻣﺒﻨﺎي اﻟﮕﻮي ﻣﺠﻤﻮﻋﻪ داده ﻫﺎي ﻣﮑﺮر، 170 درﺧﻮاﺳﺖ ﭘﺮﺗﮑﺮار از ﻣﯿﺎن ﮐﻞ دﯾﺘﺎﺳﺖﻫﺎي ﻣﻮﺟﻮد اﺳﺘﺨﺮاج ﺷﺪ. ﻧﺘﺎﯾﺞ از ﻃﺮﯾﻖ ﭘﯿﺎدهﺳﺎزي اﯾﻦ دادهﻫﺎي ﭘﺮﺗﮑﺮار در ﻧﺮم اﻓﺰار رﭘﯿﺪ ﻣﺎﯾﻨﺮ، ﺑﻪ دﺳﺖ آﻣﺪ. ﯾﺎﻓﺘﻪ ﻫﺎ: ﺑﻬﺘﺮﯾﻦ ﺣﺎﻟﺖ دﺳﺘﺮﺳﯽ ﺑﻪ ﻣﺠﻤﻮﻋﻪ دادهﻫﺎي ﺗﻮﻟﯿﺪ ﺷﺪه، ﺑﺎ ﺗﻨﻈﯿﻢ درﺟﻪ ﺣﺪاﻗﻞ ﭘﺸﺘﯿﺒﺎﻧﯽ 0/02 و ﺣﺪاﻗﻞ اﻃﻤﯿﻨﺎن 0/095 اﺳﺖ ﮐﻪ ﻣﻨﺠﺮ ﺑﻪ ﮐﺸﻒ اﻟﮕﻮرﯾﺘﻢﻫﺎي ﺷﺮﻃﯽ )اﮔﺮ – آﻧﮕﺎه( ﮔﺮدﯾﺪ. اﯾﻦ ﯾﺎﻓﺘﻪ ﻧﺸﺎن ﻣﯽدﻫﺪ: اﮔﺮ ﮐﺎرﺑﺮي در زﻣﺎن ﺟﺴﺘﺠﻮي ﻧﺮماﻓﺰار ﮐﺘﺎﺑﺨﺎﻧﻪ دﯾﺠﯿﺘﺎل ﻣﻮﺿﻮﻋﺎﺗﯽ ﻫﻤﭽﻮن )ﻋﻠﻢ اﺻﻮل، اﺟﺘﻬﺎد، ﺳﻨﺖ و...( را اﻧﺘﺨﺎب ﻧﻤﺎﯾﺪ، ﺑﻪ دﻟﯿﻞ ﺳﺎﺑﻘﻪ ﺟﺴﺘﺠﻮﻫﺎي ﭘﺮ ﺗﮑﺮار ﮐﺎرﺑﺮان ﭘﯿﺸﯿﻦ ﺑﺎ ﻫﻤﯿﻦ ﻣﻮﺿﻮﻋﺎت، از ﺳﻮي ﺳﺎﻣﺎﻧﻪ ﺗﻮﺻﯿﻪﮔﺮ، آﻧﮕﺎه ﻋﻨﺎوﯾﻦ ﻣﺮﺗﺒﻄﯽ ﺑﺎ ﻣﻮﺿﻮع اﺻﻮل ﻓﻘﻪ ﭘﯿﺸﻨﻬﺎد ﺧﻮاﻫﺪ ﺷﺪ. ﻧﺘﺎﯾﺞ: ﻃﺮاﺣﯽ ﻣﺪل ﺳﺎﻣﺎﻧﻪ ﺗﻮﺻﯿﻪﮔﺮ و اﺟﺮاي آن، ﻣﻨﺠﺮ ﺑﻪ اﻓﺰاﯾﺶ آﮔﺎﻫﯽ و ﻋﺪم ﺳﺮدرﮔﻤﯽ ﮐﺎرﺑﺮان و ارﺗﻘﺎء ﺳﻄﺢ ﮐﯿﻔﯿﺖ ﺧﺪﻣﺎت و رﺿﺎﯾﺖ آنﻫﺎ و در ﻧﻬﺎﯾﺖ ﺣﻔﻆ ﮐﺎرﺑﺮان ﻓﻌﻠﯽ و ﺟﺬب ﮐﺎرﺑﺮان ﺟﺪﯾﺪ ﺧﻮاﻫﺪ ﺷﺪ.

Objective: This research aims to present the design of the applied model in the field of personalization Thechnology of the Digital Library of Astan-e Quds-e Razavi based on the Basket Analysis pattern (association rules) and the FP-Growth algorithm regarding data mining, obtained through the implementation of Frequent itemset of users in the RapidMiner software Creating such a desirable recommendation system can provide suggestions to improve the retrieval of related information resources interesting to users in order to provide useful services and increase their request from this website. Methodology: This research is an applied research carried out using the association rules and FP-Growth wizardry which is a superior type of the issues raised in the data mining tool. The research community has 960 users of the Digital Library of Astan-e Quds-e Razavi in two-years among which based on the Frequent pattern, 170 Frequent requests were extracted from all the final data sets of document delivery service module in digital library. Other tools for accessing the Basket Analysis pattern are the utilization of RapidMiner software through which with implementation rules of data and Launchs association rules operators and the FP-Growth algorithm, and more importantly, a change in the minimum degree of support and confidence, lead to The production of new association rules, which can be interpreted by these rules, offering a suggestion for the design of an appropriate model for the recommendation system in the digital library website. Findings: The results showed that the best access to generated rules by setting the minimum support of 0.02 and a minimum confidence of 0.095 resulted in the creation of 1081 new rules, indicating if the user website searches for topics in “Osole Fiqh” such as (osole Amalieh, Ijtihad and Quran), because of the frequent searching records of the previous users whith the same subject matter, that the recommender system paves the way for suggesting& gaining access to “Risalah Amaliyah” with recorded number (2309) along with his main search, the manuscript “Zavabetol Osul” with recorded number (38696) finds amoung his/her searches. Therefore, all of the following subset of “Uṣūl al-fiqh” can be predicted through the analysis of association rules and provide recommendations to the users of the digital library system to search for effective and relevant subjectes relative to the users requests in similar topics & titley. Conclusion: Since the generation of personalization technology and its implementation on the website of digital libraries in the form of recommendation system is based on the establishment of interaction between users and the modern services of digital libraries in form of user interface effective, this technology brings about increasing specialized knowledge, lack of users wondering improving the quality of service and user satisfaction, and ultimately creating value added for libraries. Therefore, with such an approach, dealing with modern services in the form of customer-service delivery, is of the most importance in digital libraries.