The application of extension data mining in behavior type guide residential door model design

Размещено на http://www.allbest.ru/

School of Architecture, Harbin Institute of Technology, Harbin, PRC

School of Architecture, Jilin Jianzhu University

The application of extension data mining in behavior type guide residential door model design

Zhihui Gao,

Lecturer, Changchun, PRC; postgraduate student,

Abstract

This article from the perspective of cross-disciplinary, select a residential district as the research object in Changchun, a residential area residents to behavior type guide residential door model design, through field investigation, questionnaire and the extension data mining methods, excavated in the law of the behavior and characteristics, and analysis of the factors of influence in this behavior, behavior characteristics and living within grasp their house demand, on the basis of summed up the corresponding residential door model design. This article research goal lies in the limited living space, improve efficiency room, dug up to meet the current people's behavior patterns within the residential door model design. And the method of extension data mining first applies to the residential indoor behavior pattern research, so as to provide an objective basis for the future residential construction design and practice reference.

Keywords: behavior guide type; Door model design; Extensive data mining; The service efficiency of the room.

Абстракт

ПРИМЕНЕНИЕ РАСШИРЕННОГО АНАЛИЗА ДАННЫХ В ПРОЕКТИРОВАНИИ МОДЕЛЕЙ ДВЕРЕЙ С УЧЕТОМ ТИПА ИСПОЛЬЗОВАНИЯ

Чжихуэй Гао,

преподаватель школы архитектуры, Цзилиньский строительный университет, Чанчунь, КНР; аспирант школы архитектуры, Харбинский политехнический университет, Харбин, КНР;

Эта статья с точки зрения междисциплинарности, основана на выборе жилого района в городе Чанчун в качестве объекта исследования, жители жилого района заинтересованы в разработке проекта моделей жилых дверей, который проводится авторами посредством исследований, анкетирования и расширения данных о методах добычи полезных ископаемых, также на основании закономерностей в поведении и анализе факторов влияния на поведение, также учитывается спрос людей, проживающих в этом районе. На основе этих данных спроектирован дизайн модели жилой двери. Цель исследования этой статьи - в условиях ограниченности жилого пространства увеличить процент полезной площади в комнате. Расширенный метод изучения данных впервые применяется в исследовании модели поведения людей в жилом помещении, он обеспечивает объективную основу для будущего проектирования жилого строительства и ориентацию для практического применения в дальнейшем.

Ключевые слова: поведенческий тип руководства; дизайн модели двери; расширенный поиск данных; эффективность обслуживания помещения

Introduction

For residential buildings, the door model is the important measure of the good or bad, a significant role in the housing design. Usually family has the following a few big functional spaces: living room, dining room, kitchen, bedroom, bathroom, storeroom, etc. The trade-off with the combination of function space, meet the needs of different plane model f Can choose a family, to a large extent determines the habitant, the behavior way of life, on the other hand, the behavior of habitant characteristics will also directly to verify the door model, both are interdependent, inseparable.

Previous door model design, architects are mostly based on experience and intuition on the basis of personal preference and subjective, its way of thinking and logic basis relies on the burst of inspiration, this is not conducive to residential door model design on the path of scientific and intelligent development. Exampled by a certain residential area of Changchun city residents' behavior characteristics as the research object, by means of extension data mining method to classify indoor behavior data research, attempts to residents in the relationship between behavior and door model design for an empirical study, in order to analysis and solve the conflict problems between behavior characteristic and the door model design.

2. Analysis of extension data mining in behavior guide type door model design.

2.1. Analysis of the extension data mining. Extension data mining 2-4 is the product of extenics is combined with data mining, it studies using the theory and method of extenic to mining in the database or knowledge base and knowledge transformation to solve conflict the problem 5. Conflict problem is that means can not be achieved under the existing conditions for the goal of the problem. To solve the conflict problem must first clear the heart of the problem, through to the target and the condition transformation to achieve the purpose to solve the problem of contradiction. Extension data mining based on primitive logic cells on the method of formal system combined with database and data warehouse, data form suitable for the "transformation" of knowledge representation method 3. Usually the extension set is divided into five fields: the quality change domain of positive and negative, quantity change domain of positive and negative, the expand domain. the occurrence of quantity exceeds the critical value will lead to qualitative change, reaches the critical value is called extension, is the basis of the five domains are interdependent, also is the core content of classification. Using the quantitative and qualitative change single transform or transform into contradiction problems, finally achieve the purpose to solve the problem.

Analysis of behavior guide type door model design

This article advocated behavior guide type door model design is centered on residents, based on the indoor behavior as the starting point of habitant door model design, with the actual behavior of habitant characteristics as the main body, residential door model study. Using the method of extension data mining, digging out the habitant indoor behaviors and characteristics, on the basis of summed up the corresponding residential door model design. residential door room

Analysis of the procedures of extension data mining for behavior guide type door model design

Establish behavior guide type door model design the core issue of the matter-element model. Extension data mining and data mining, first need to define the target question 5. In guide type with indoor behavior as the breakthrough point of the research in the process of solving the problem of family, is the core issue of research in the limited living space, improve efficiency room, and dug up to meet the current people's behavior patterns within the residential door model design. Core problem, expressed in P; L represent the condition of existing information, G represents achieve goals, you need to solve the problem to be expressed as:

Choose door model the design assessment information. Have data about the door model design are often very large and complex, such as family members, indoor behavior of streamline, economic capacity, rooms, etc. only their influence degree of different door model design. Selection based on this, the nuclear issue of the indoor room use efficiency, streamline evaluation characteristics as indoor behavior. In this paper, the specific studies in this room to use time, streamline and efficient of the room, as the indoor behavior data. such as with D on behalf of the assessment information; D, on behalf of the evaluation information, u representative element numerical evaluation information, they will evaluate the information presented as:

Establish evaluation information related database. Selected evaluation information can determine the database and evaluation information related to the original information. Some characteristics of evaluation can be obtained directly from the database, such as indoor behavioral streamline; Others cannot be directly obtained from the database, need through the other indirect data and relational database based on extension data mining, such as indoor room efficiency. With I on behalf of the original information related to evaluation assessment information, c on behalf of the original information, v on behalf of the original information of value, will evaluate the information and the corresponding original expressed as:

Behavior patterns mining and rule generation. Use appropriate mining algorithm, the core issue of contradiction problem, digging out the indoor behavior data implied, regularity of knowledge, and knowledge as analysis present door model design, and adopt corresponding change on the strategies.

Examples for extension data mining in behavior guide type door model design. On the basis of comprehensive investigation changchun city mainly residential area, is located in the selected area Jingyue district in a residential area as the research object, the area of education, tourism, health, business and other supporting facilities. Residential area between 65 ~ 160 m2, from one room to the four rooms, meet a variety of door model analysis. Investigation on April 25th to 27th, 2014, and on May 1st ~ 2nd, each 10 households in the selected four families, methods such as questionnaires, user description record indoor behavior in order to get the initial data. Including singles 13, the core families 19 households, six main households and other two class families. Basic information questionnaire mainly relates to the basic information, indoor behavior and indoor problems three aspects.

Established family problems of the matter-element model. In indoor behavior patterns dominated door model the design as the breakthrough point of the research in the process of solving the problem of family, is the core issue of research is in the limited living space(L), designed to meet all people's behavior(G) in family function space. Door model design problems of matter-element model is as follow:

Selected behavior guide door model the design of evaluation information.

The basic information questionnaire, age, cultural level, family population composition and indoor each behavior classification analysis one by one. Based on the core problem of consideration, this article selects the efficiency of indoor room, indoor behavior streamline as evaluation characteristics. Behavior leading type door model design corresponding evaluation information specific problem are as follows:

3.3 Establish relevant database. Streamline indoor behavior evaluation features can be directly obtained through the research; While indoor room efficiency cannot be directly obtained from the database, need through a single room with time and surveyed the room total used time, to get the corresponding evaluation information, expressed the original information is as follows:

Room use time, for example, starting from indoor behavior primitives, the "behavior", " behavioral characteristic " and "behavioral characteristic value" extension divergence analysis it is concluded that the following several factors, expressed as:

At the same time, the family members can be further spread for each family demographic data. Requirements can be determined by the related database, the design of the questionnaire mainly involves the user basic information, indoor behavior, satisfaction survey in three aspects. Basic information including gender, age, educational level, family of three events; Indoor behaviors of family members to use each room's time and behavioral streamline; Satisfaction survey respectively research family members for indoor evaluation information between each room.

3.4 Mining and rule generation. In the investigation of all the family, family b and family c were most widely used, generality, indicates that demand for these two kinds of family is the most intense. By comparison the basic information for the family and indoor behavior, the results found that family of population is the biggest variable combination and family influence within the family behavior, culture degree. This means family population and culture degree in maximum extent affects indoor behavior and family composition.

Figure 1: Frame of behavior guide type door model design

Through the statistical analysis of survey data and mining, throughout each time the use efficiency of indoor room (fig. 2). Due to the toilet, locker room use time of uncertainty, and to enhance the accuracy of the research, to each room general use time expression as the main rooms of the general used of time, for the living room, bedroom, study, restaurant total used time.

Take the user description method of four family indoor behavior of streamline, indoor behavior resulting from the superposition of chart (fig. 3). Line represents the degree of frequent use of the route of thickness (fig. 3).

Figure 2. Superposition of different area door model indoor behavioral streamline figure: a - one-bedroom60 ^70 Ш; b - two-bedroom70 ^90 Ш; c - three-bedroom90~120 Ш2;

d - four-bedroom140^160 Ш

By the two diagrams can be found that although the bedroom is indoor used the longest room, has obvious time zone boundaries, use frequency is highest, night activities and behavior of a single, used only for rest; For one and two bedrooms', cultural degree is higher, the higher the efficiency of the bedroom, because most of the bedroom a single functional space composite used for "rest and study". Three and four bedrooms' members used frequency is not obvious. In terms of streamline, due to the function rooms, where a longer results in a streamline in the home also. In the satisfaction survey, also confirm the residents demand for dismantling composite function room. Single from the point of use efficiency, with the increase of family room, bedroom use efficiency gradually reduce, in the big door model, designers can consider to change the bedrooms' used properties (qualitative) to improve the efficiency of using the method, such as the spare room for storeroom or gym, etc.

Living room is the second time the longest room, in the four-bedroom's door model, the function of the living rooms are the most frequent activities rooms, occupied the important positions in the door model design. Average used time for 4 hours, at 8and16 o' clock reach the peak in the frequency, and actively also, no time limited; As the increase of family composition, to gradually increased the used frequency of living room (quantity); The trunk for the maximum efficiency of the function room, almost for older people to use, and almost no cross with their children.

The average use time of dining room is 2.5 hours, and the frequency of peak and the living room just complementary; Show the living room and dining room under the condition of limited use in indoor area, stacks setting (qualitative).

For one of the bedroom changed into a study room's (qualitative) family, the study room's average used time is 3.5 hours, the study is usually accompanied by a peak and coming after the use frequency of living room. And with the improvement of family culture degree used increasingly frequency.

3.5 The results and discussion. Behavior guide type door model design change to designers as the main body of the design principle in the past, main factors that influence the residents and their behavior as, return to the origin in residential door model design, based on fully embody the people-oriented design idea. In addition, this study only from the efficiency of room and indoor behavior streamline two assessment information sets to investigate the relationship between the behavior and door model design, with the deepening of the research and the needs of designers, can according to need to determine the assessment information and its corresponding number and variety of the original information, dig deeper, implicit, regularity of knowledge; As long as the suitable mining algorithm, combined with the core issue of contradiction problem, can according to the design bottleneck changed; Dig up the regularity of knowledge can provides the policy basis for the present door model design.

Conclusion

Extension data mining methods as a cross subject, cooperate with relevant questionnaire data acquisition mode, such as residential building design for comprehensive analysis of family properties, and solve the contradictions of the behavior guidance door model design, provides a new design perspective, has set up a bridge for the rational analysis. At this point, the designers can not only make qualitative and quantitative analysis on door model design, can also according to the laws of the excavated implied information and take corresponding measures to make door model design conforms to the behavior characteristics of family members. In addition, another contribution of extension data mining method is to translate the data using the extension model multiple structured language, unified the formalized in the form of visual expression, will become the new development direction for the future design.

References

1. Yanmin Zhou, etc. Residential fine design. Beijing: China building industry press, 2008.

2. Chunyan Yang, Wen Cai. Extension engineering. Beijing: science press, 2007.

3. Wen Cai, Chunyan Yang, Wenwei Chen, Xingsen Li. Extension set and extension data mining. Beijing: science press, 2008.

4. Wenwei Chen. Mining change knowledge extension data mining research. China engineering science. 2006, 11:70-73.

5. Chunyan Yang, Xiaomei Li, Wenwei Chen, Wen Cai. Extension data mining method and its computer implementation. Guangzhou: guangzhou higher education press, 2010.

6. Sun Yuan. Commercial housing residential door model design. City building. 2014, 2:22.

7. Lu-min Wang, Junping Xu. In this behavior patterns and collection of residential landscape - China prevue multi-family segment change since 1949. The new building, 2003, 6:34 ~ 36.

Размещено на Allbest.ru