Socialization rate
The socialization ratio (CS) is a value that a social network user automatically acquires. This ratio shows the level of integration of a specific user into the social network.
The COP is, in most cases, only an indicator, so it can not affect the size of other values ​​that characterize the user. A CS may include various objective mechanical and dependent rating numerical values, the effect of other social networks and other parameters.
The COP consists of the following indicators:
This characteristic allows you to define a circle of friends of the user and describe them with the help of logically significant units - tags . Each friend is given a logical load with which it relates to the user. The load can be expressed as: colleague, comrade, classmate, friend, relative. In the tags you will not see the word "enemy", since, as a rule, users do not add people to the list of people to whom they feel a certain amount of dislike. Using the description of each circle of a user’s friends, one can define his own characteristic in terms of relations with these users and determine the degree of his load on his friends (in fact, calculate the coefficient of user interaction with his contacts). This can be done mechanically, but, in any case, such a process will be based on the desire of the user himself: he himself associates a specific tag (s) with a specific user.
Such links can be presented in the form of links to the profiles of friends on the social network using XFN, or other means of indicating the relativity of the relationship between individuals on the Internet.
With the help of an indicator such as personalization, you can identify the various interests of the user. Personalization allows you to determine the range of interests of the user, after which it will be possible to participate in helping to find new friends who have exactly the same interests and preferences. Based on this strategy, there are already a sufficient number of narrowly focused thematic networks (Last.fm and others). On such resources, it is often possible to find various indicators of compatibility between different individuals. Very often, these services provide an API for exporting a variety of information that is closely related to a specific user.
Despite the fact that the rating values ​​are subjective and depend on the goals pursued by the resource authors (clarification of certain indicators of relationships, assessment, credibility, etc.), they can be used to form an opinion about the user, assessment and user characteristics.
Ratings can be of several types:
All presented types of ratings can already be found in many social networks. In order to calculate the CS, using, among other things, this indicator, it is necessary to develop an individual approach to indexing a specific rating on a specific social network or other type of resource, since, I repeat, the algorithms and goals of each indicator differ.
Profiling allows you to gather information about a person first-hand, that is, from the person herself. Relying on such data in today's Internet is not worth much, as the data written about yourself is subjective (and in many cases even falsified). But you shouldn’t neglect them either, because you can find out the basic characteristics of the user: age category, gender, manually formed list of interests. In addition, profiling with mechanical (manual) analysis allows for the psychological assessment of the user based on the style of writing the text, manner of speaking and other factors.
Personal information will help to form a view of the professional skills and abilities of the user, in addition, such data is easily verified. Such information is very often not needed by other users; therefore, it is considered official and is only needed for the personnel policy regarding the person.
Perhaps one of the most important resources that describes the relationship of society (community) to the user. The citation index allows you to determine the general attitude of society to the user. That is, a negative index will indicate that the user has made a negative impression on specific participants in the society. A positive index, by contrast, indicates a good attitude towards the user. From this we can conclude that the user enters the network with a zero citation index (and other characteristics will be relative to this figure, if this person obviously did not meet elsewhere).
Calculation of such a magnitude as a COP is difficult to imagine in the form of an automated process, as a network of social networks and narrow thematic resources are growing at an enormous pace, and not all of them provide API / direct access to information regarding the personalization of their users. But other indicators can be calculated with the assistance of third-party forces, such as search engines, semantic analysis systems, as well as the full assistance of the user himself: the provision of a URI for his accounts on social networks and other resources, assistance in calculating person indexation, person quoting, and a description of his own activity online.
It is worth noting that the majority of indices in the CS can take positive and negative values, which clearly expresses one or another characteristic of the user. The index sign, "+" or "-", indicates the appropriate relationship to the person.
For the zero mark of all indices is taken the number zero. “0” is the basis of all calculations and the initial socialization factor of the user in the network.
It is also necessary to distinguish the scope of application of the CS: this value can be the relative size of the areas of calculation and analysis taken for calculation. Thus, in one social group, a user may be in good standing, but on the whole in a social network he may also be at a level below the average. But it should also be understood that there may be several groups on the network at one level of correlation, but in one of them a specific user will receive KS +, and in the other will already receive a negative KS. Therefore, the analysis with respect to single-stage groups and bundles should be neglected, since for the most part it is not an effective assessment of the user’s CS.
Analysis and development of algorithms for computing a CS can be done indefinitely, but I think this is worth doing. Already, one can see the intersection of this concept with the concept of a semantic network, which means that, by studying socialization, one can learn new things and make inferences about semantic networks. And the introduced concept is to a greater extent a statistical value, but despite this it shows the real position of the user in the social network.
This article is a small analytical collection of the author’s own thoughts.
The COP is, in most cases, only an indicator, so it can not affect the size of other values ​​that characterize the user. A CS may include various objective mechanical and dependent rating numerical values, the effect of other social networks and other parameters.
Components
The COP consists of the following indicators:
- Coverage of messaging services and contacts This includes user activity in various IM services, the number of contacts adjacent to it (contacts that the user somehow recognized by adding him to his white list; the concept is rather subjective, since the reasons for adding a user to white and black lists may be unmotivated or poorly motivated).
') - Personalization by using services with a narrow thematic focus. This item defines thematic preferences of the user (music, books, films, programs, etc.). Based on the data of thematic preferences, as well as the very fact of participation in such social services (person socialization services), it is possible to collect a lot of statistical information about a specific user.
- Mechanical and automatic (dependent) rating numerical values. A similar type of indicator can strongly influence the socialization rate, but in fact it is weakly associated with previous indicators, since the calculation of different numerical values ​​is based on completely different algorithms and pursue different goals, therefore the objectivity of a given The indicator may be questioned (which will require additional research in this direction).
- Profiling and providing personal information With the help of providing personal information, you can create not only a picture of your personal views and needs, but also portray your professional portrait using such means as a resume, questionnaire, etc. Person profiling has long been used on the Internet and is requested when a user is registered or after confirming the validity of his account. Profiling includes the user's personal data (gender, age, interests, preferences; depending on the type of resource on which the registration is taking place). The objectivity and veracity of the data submitted by the user is a very weak indicator, since he fills in the user’s profile himself, so you should not trust such data without additional verification.
- Indexation of the personThe indexing element allows to assess the degree of integration of the person in the network using existing search tools and tools for analyzing the status of the bank of requests / documents on the Internet. Indexing a person is the level of citing him on the Internet (using a real name or a pseudonym of a person on resources on the network, except for the resource of the user himself, quoting his articles, links to materials written by this user, reviews on the user by other network participants, etc.) . Indexation of a person is one of the most powerful indicators of socialization of a person, and also determines the extent of his media coverage on the Internet. Two other indicators follow from person indexation : person citation indexation and person activity level in the network.
Coverage of messaging services and contacts
This characteristic allows you to define a circle of friends of the user and describe them with the help of logically significant units - tags . Each friend is given a logical load with which it relates to the user. The load can be expressed as: colleague, comrade, classmate, friend, relative. In the tags you will not see the word "enemy", since, as a rule, users do not add people to the list of people to whom they feel a certain amount of dislike. Using the description of each circle of a user’s friends, one can define his own characteristic in terms of relations with these users and determine the degree of his load on his friends (in fact, calculate the coefficient of user interaction with his contacts). This can be done mechanically, but, in any case, such a process will be based on the desire of the user himself: he himself associates a specific tag (s) with a specific user.
Such links can be presented in the form of links to the profiles of friends on the social network using XFN, or other means of indicating the relativity of the relationship between individuals on the Internet.
Personalization and thematic preferences
With the help of an indicator such as personalization, you can identify the various interests of the user. Personalization allows you to determine the range of interests of the user, after which it will be possible to participate in helping to find new friends who have exactly the same interests and preferences. Based on this strategy, there are already a sufficient number of narrowly focused thematic networks (Last.fm and others). On such resources, it is often possible to find various indicators of compatibility between different individuals. Very often, these services provide an API for exporting a variety of information that is closely related to a specific user.
Rating Numerical Values
Despite the fact that the rating values ​​are subjective and depend on the goals pursued by the resource authors (clarification of certain indicators of relationships, assessment, credibility, etc.), they can be used to form an opinion about the user, assessment and user characteristics.
Ratings can be of several types:
- Rating assessment of the user’s action (action implies material laid out by the user, or an event involving the latter)
- User Rating and Authority Rating
- Compatibility Indicator for any surrounding network object
All presented types of ratings can already be found in many social networks. In order to calculate the CS, using, among other things, this indicator, it is necessary to develop an individual approach to indexing a specific rating on a specific social network or other type of resource, since, I repeat, the algorithms and goals of each indicator differ.
Profiling
Profiling allows you to gather information about a person first-hand, that is, from the person herself. Relying on such data in today's Internet is not worth much, as the data written about yourself is subjective (and in many cases even falsified). But you shouldn’t neglect them either, because you can find out the basic characteristics of the user: age category, gender, manually formed list of interests. In addition, profiling with mechanical (manual) analysis allows for the psychological assessment of the user based on the style of writing the text, manner of speaking and other factors.
Provision of personal information
Personal information will help to form a view of the professional skills and abilities of the user, in addition, such data is easily verified. Such information is very often not needed by other users; therefore, it is considered official and is only needed for the personnel policy regarding the person.
Person indexing
Perhaps one of the most important resources that describes the relationship of society (community) to the user. The citation index allows you to determine the general attitude of society to the user. That is, a negative index will indicate that the user has made a negative impression on specific participants in the society. A positive index, by contrast, indicates a good attitude towards the user. From this we can conclude that the user enters the network with a zero citation index (and other characteristics will be relative to this figure, if this person obviously did not meet elsewhere).
Payment
Calculation of such a magnitude as a COP is difficult to imagine in the form of an automated process, as a network of social networks and narrow thematic resources are growing at an enormous pace, and not all of them provide API / direct access to information regarding the personalization of their users. But other indicators can be calculated with the assistance of third-party forces, such as search engines, semantic analysis systems, as well as the full assistance of the user himself: the provision of a URI for his accounts on social networks and other resources, assistance in calculating person indexation, person quoting, and a description of his own activity online.
It is worth noting that the majority of indices in the CS can take positive and negative values, which clearly expresses one or another characteristic of the user. The index sign, "+" or "-", indicates the appropriate relationship to the person.
For the zero mark of all indices is taken the number zero. “0” is the basis of all calculations and the initial socialization factor of the user in the network.
It is also necessary to distinguish the scope of application of the CS: this value can be the relative size of the areas of calculation and analysis taken for calculation. Thus, in one social group, a user may be in good standing, but on the whole in a social network he may also be at a level below the average. But it should also be understood that there may be several groups on the network at one level of correlation, but in one of them a specific user will receive KS +, and in the other will already receive a negative KS. Therefore, the analysis with respect to single-stage groups and bundles should be neglected, since for the most part it is not an effective assessment of the user’s CS.
Subtotals
Analysis and development of algorithms for computing a CS can be done indefinitely, but I think this is worth doing. Already, one can see the intersection of this concept with the concept of a semantic network, which means that, by studying socialization, one can learn new things and make inferences about semantic networks. And the introduced concept is to a greater extent a statistical value, but despite this it shows the real position of the user in the social network.
This article is a small analytical collection of the author’s own thoughts.
Source: https://habr.com/ru/post/21421/
All Articles