Skip to content

petchat/senz.sdk.android.maven

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

88 Commits
 
 
 
 

Repository files navigation

senz-sdk-android

1、 Android studio 中引用

在工程目录 build.gradle 文件中添加如下内容:

buildscript {
    repositories {
    jcenter()
    maven {
                url "https://raw.githubusercontent.com/petchat/senz.sdk.android.maven/master"
        }
        maven { 
                url "https://jitpack.io" 
                }
    }
}

allprojects {
    repositories {
        jcenter()
        maven {
            url "https://raw.githubusercontent.com/petchat/senz.sdk.android.maven/master"
        }
        maven { url "https://jitpack.io" }
    }
}

在引用模块下的build.gradle (minSdkVersion >=15)

android {
    packagingOptions {
        exclude 'META-INF/LICENSE'
        exclude 'META-INF/NOTICE'
    }
    dependencies {
        compile 'io.petchat:senzsdk:2.0.17'
        compile 'com.wilddog:wilddog-client-android:0.5.1+'
            compile 'com.github.zishell:MotionClassifyLib:v1.0.5'
            
    }
}

2、 使用SENZ SDK

2.1 初始化

在MainActivity 的 onCreate()中(必须在UI主线程中调用),调用:

//please register on senz server to get your own appId
    Senz.initialize(MainActivity.this,appId);

appId为您在Dashboard注册新建的应用id。

2.2 Senz核心api接口

####2.2.1 用户属性UserInfo类 该类主要用于分析用户的属性,如:年龄、性别、消费能力、社会属性等。

调用例子:

    UserInfo userInfo = new UserInfo(this);
        userInfo.getUserInfo(new SenzCallback() {
            @Override
            public void done(Exception e, Object object) {
                if (e == null) {
                    HashMap<String, Object> map = (HashMap<String, Object>) object;
                    //...
                   
                } else {
                   //...
                }
            }
        });

#####hashMap数据格式说明 UserInfo 包含用户属性和用户喜好, 用户属性:

{
    "age": [
        "16down",
        "16to35",
        "35to55",
        "55up"
    ],
    "has_pet": [
        "positive",
        "negative"
    ],
    "has_car": [
        "positive",
        "negative"
    ],
    "occupation": [
        "engineer",
        "salesman",
        "teacher",
        "student",
        "soldier",
        "official",
        "supervisor",
        "freelancer",
        "others"
    ],
    "pregnant": [
        "positive",
        "negative"
    ],
    "field": [
        "infotech",
        "commerce",
        "law",
        "athlete",
        "medical",
        "human_resource",
        "financial",
        "architecture",
        "humanities",
        "natural",
        "manufacture",
        "agriculture",
        "service"
    ],
    "marriage": [
        "positive",
        "negative"
    ],
    "consumption": [
        "5000down",
        "5000to10000",
        "10000to20000",
        "20000up"
    ],
    "gender": [
        "positive",
        "negative"
    ]
}

用户喜好:

{
    "sport": [
        "jogging",
        "fitness",
        "basketball",
        "football",
        "badminton",
        "bicycling",
        "table_tennis"
    ],
    "sports_news": [
        "positive",
        "negative"
    ],
    "variety_show": [
        "positive",
        "negative"
    ],
    "tvseries_show": [
        "positive",
        "negative"
    ],
    "study": [
        "positive",
        "negative"
    ],
    "social": [
        "positive",
        "negative"
    ],
    "offline_shoppng": [
        "positive",
        "negative"
    ],
    "indoorsman": [
        "positive",
        "negative"
    ],
    "gamer": [
        "positive",
        "negative"
    ],
    "sports_show": [
        "positive",
        "negative"
    ],
    "online_shopping": [
        "positive",
        "negative"
    ],
    "current_news": [
        "positive",
        "negative"
    ],
    "business_news": [
        "positive",
        "negative"
    ],
    "game_news": [
        "positive",
        "negative"
    ],
    "health": [
        "positive",
        "negative"
    ],
    "game_show": [
        "positive",
        "negative"
    ],
    "acg": [
        "positive",
        "negative"
    ],
    "tech_news": [
        "positive",
        "negative"
    ],
    "entertainment_news": [
        "positive",
        "negative"
    ]
}

例如,如果您想获取用户的消费信息:

if (map.containsKey("consumption")) {
        HashMap<String, Double> consumption = (HashMap<String, Double>) map.get("consumption");
        }

返回一个HashMap<String, Double>:

10000to20000 : 0.24214
20000up : 0.15570
5000down : 0.15272
5000to10000 : 0.13414

如果想获取用户在tech_news的喜好,则返回一个Double值。即,如果key下有子类会返回HashMap,如果没有子类则返回相应的预测值(double)

####2.2.2 用户的情境识别 context(deprecated)

调用例子:

    UserContext userContext = new UserContext(this);
       userContext.getUserContext(new SenzCallback() {
           @Override
           public void done(Exception e, Object object) {
               if (e == null) {
                   SenzContext senzContext = (SenzContext) object;
                   //...
               } else {
                   //...
               }
           }
       });

#####数据格式说明 SenzContext类:

public class SenzContext {
    public HashMap<String, Double> poiProbLv1;
    public HashMap<String, Double> poiProbLv2;
    public HashMap<String, Double> motionProb;
    public HashMap<String, Double> sceneProb;
    public String locationUpdatedAt;
    public String motionUpdatedAt;
    public String sceneUpdatedAt;
    }

poiProbLv1 和 poiProbLv2 为用户location的情境识别, sceneProb 为用户的所处场景识别, motionProb 为 用户的动作情境识别。 属性保存在了key中,对应的值保存在了value中, 键值不是固定的,请使用者自行遍历相应的map。

####2.2.3 用户事件识别 event event指用户发生的事件,如某个时间段,用户在看电影或者逛商场。 event所包含的所有事件类型:

0. 在家: contextAtHome
1. 上班路上: contextCommutingWork
2. 在公司: contextAtWork
3. 回家路上: contextCommutingHome
4. 商圈工作中: contextWorkingInCBD
5. 学校上课中: contextStudyingInSchool {mapping to contextAtSchool.png}
6. 学校工作中: contextWorkingInSchool {mapping to contextAtSchool.png}
7. 户外锻炼: contextOutdoorExercise
8. 室内锻炼 contextIndoorExercise
9. 在餐厅吃饭: contextDinningOut
10. 旅游: contextTravelling
11. 出行: contextGoingOut
12. 聚会: contextInParty
13. 逛街: contextWindowShopping 
14. 看电影: contextAtCinema
15. 展览会: contextAtExhibition
16. 演唱会: contextAtPopsConcert
17. 戏剧: contextAtTheatre
18. 音乐会: contextAtClassicsConcert
19. 在家休息:contextRelaxing

event值的获取需要注册监听:

action: senz.intent.action.EVENT

调用例子:

       BroadcastReceiver eventReceiver = new BroadcastReceiver() {
       @Override
       public void onReceive(Context context, Intent intent) {
           if ("senz.intent.action.EVENT".equals(intent.getAction())) {
               Bundle bundle = intent.getExtras();
               Serializable data = bundle.getSerializable("event");
               SenzEvent event = (SenzEvent) data;
               //...
           }
       }
   };

#####数据格式说明 SenzEvent 类:

public class SenzEvent implements Serializable {
   public long timestamp;     //更新时间
   public String eventType;   //event类型
   public double probability;  //可信度
   }

####2.2.4 用户的 home 和 office 的识别和监听 senz可以自动的识别用户的家庭和工作地点,也可以检测出用户离开、进入家或者公司。 检测出这些变化时,senz会发出相应的广播。

监听action:

action: senz.intent.action.HOME_OFFICE_STATUS

调用例子:

       BroadcastReceiver statusReceiver = new BroadcastReceiver() {
       @Override
       public void onReceive(Context context, Intent intent) {
           if ("senz.intent.action.HOME_OFFICE_STATUS".equals(intent.getAction())) {
               Bundle bundle = intent.getExtras();
               Serializable data = bundle.getSerializable("home_office_status");
               SenzOHStatus status = (SenzOHStatus) data;
               //...
           }
       }
   };

#####数据格式说明 SenzOHStatus 类:

public class SenzOHStatus implements Serializable {
   public long updateTime;   //更新时间
   public String stautsType;  //状态类型
   public double probability;  //可信度
   }

####2.2.5 用户 motion 的识别和 motion 改变的监听

senz可以识别出用户当前的动作,如 坐、跑步、走路等。当检测到用户的动作发生改变时 senz会发出相应的广播。

监听action:

action: senz.intent.action.MOTION_CHANGED

调用例子:

       BroadcastReceiver motionReceiver = new BroadcastReceiver() {
       @Override
       public void onReceive(Context context, Intent intent) {
           if ("senz.intent.action.MOTION_CHANGED".equals(intent.getAction())) {
               Bundle bundle = intent.getExtras();
               Serializable data = bundle.getSerializable("motion_changed");
               SenzMotion motion = (SenzMotion) data;
               //...
           }
       }
   };

#####数据格式说明 SenzMotion 类:

public class SenzMotion implements Serializable {
   public String motionType;   //motion类型
   public double probability;   //可信度
   public long timestamp;      //更新时间
   }
motion识别的api调用
    /**
     * 如果手机没有磁场传感器,可以只用加速传感器
     * @param accRawData 加速传感器数据
     * @return
     */
 public HashMap<String, Double> getDetectMotions(final double[][] accRawData){};
 /**
     * 运用加速传感器和磁场数据提供更准确的预测结果
     * @param accRawData 加速传感器三轴数据
     * @param magRawData 磁场三轴数据
     * @return
     */
 public HashMap<String, Double> getDetectedMotions(final double[][] accRawData, final double[][] magRawData){};
 
  /**
     * TRUE:正在看手机
     * @param accRawData 加速传感器三轴数据
     * @return
     */
 public boolean isWatchPhone(final double[][] accRawData){};
 //调用例子:
 DetectMotion detectMotion = new DetectMotion(context);
 HashMap<String, Double> result = detectMotion.getDetectedMotions(accRawData, magRawData);
 
 boolean isWatchPhone = result.get("isWatchPhone") == 1.0 ? true : false;
 motionType = (int)result.get("motionType");

##2.3 设置监听的另一种方法

在manifest中注册想要监听的事件:

<receiver android:name=".YourBroadcastReceiverClassNameHere">
    <intent-filter>
        <action android:name="senz.intent.action.HOME_OFFICE_STATUS"/>
        <action android:name="senz.intent.action.MOTION_CHANGED"/>
        <action android:name="senz.intent.action.EVENT"/>
        
    </intent-filter>
</receiver>

编写自己的类,继承自 BroadcastReceiver,监听相应的action,并从intent中提取相应的值。 三种事件的值,通过序列化后的键分别为:"motion_changed", "event", "home_office_status"。 请参考下面的例子:

public class YourReceiver extends BroadcastReceiver {

    @Override
    public void onReceive(Context context, Intent intent) {
        if ("senz.intent.action.MOTION_CHANGED".equals(intent.getAction())) {
                Bundle bundle = intent.getExtras();
                Serializable data = bundle.getSerializable("motion_changed");
                SenzMotion motion = (SenzMotion) data;
                //...
            }
    }
}