In conducting a mobility analysis using Mobile Positioning Data, the most critical step is to define each customer's usual environment. The initial concept of mobility used is the movement that occurs from and to every usual environment, so errors in determining the usual environment will cause incorrect mobility statistics. Therefore, Anchor Mobility Data Analytic (AMDA) is proposed for Home-Work Location Determination from Mobile Positioning Data. This algorithm uses clockwise reversal to make it easier to classify someone in their usual environment. Unfortunately, only about 80% of the raw data can be used to establish usual environments. The remaining 20% do not have sufficient data history. This study found that the accuracy of AMDA in determining monthly home location was 98.8% at the provincial level and 88.7% at the regency level. As for the determination of monthly work locations, 98.9% at the provincial level and 70.4% at the regency level.

Until now, BPS - Statistics Indonesia has conducted monthly accommodation surveys both for the star and non-star accommodation categories to provide information on commercial accommodation activities at the national and regional levels. Both star and non-star accommodation categories are done by complete enumeration in each region. Statistics include guest night and room capacity to obtain the occupancy rate of a hotel room. The data contains daily accommodation information that is collected every month, so then it will be entered completely in each region following the observation month. Due to the timeliness requirements for monthly press releases, BPS has implemented online data entry since 2017. It may seem obvious, regions that have more interest will have an impact on a bigger number of accommodations, which also affects the number of enumerators and may lead to such problems especially in response burden. Unfortunately, the same problem is also not easily avoided by regions with less accommodation, mostly due to the distance issues to the accommodation area and its spread in the region. Therefore, a new data collection strategy is required to provide respondents with convenience in order to increase response rates, as well as to reduce the workload of enumerators which also leads to the lower cost. The outbreak of COVID-19 has posed unprecedented problems for National Statistical Offices (NSOs) around the world, including BPS – Statistics Indonesia. This crisis has led us to think in new ways and make decisions that will change our statistical operations in order to meet ongoing data needs even throughout the epidemic. The purpose of this paper is to discuss the evolution of accommodation surveys, which are designed to not only solve problems but also achieve objectives. Currently, there are nearly 180 active users of this self-enumeration accommodation survey for about 142 distinct accommodations across Indonesia. Moreover, this addition has proven to have succeeded in increasing the response rate average from 57.17% in 2020 to 68,35% in 2021.

Statistics Indonesia (BPS) has been using Mobile Positioning Data (MPD) to support official statistics since 2016. As a source of big data, MPD also has veracity characteristics, indicating uncertainty in the data. Therefore, it is necessary to check that the data are good enough to allow further analysis and the quality assurance process. Currently, there is no established international standard for quality assurance of MPD. This paper describes the quality matrix used by BPS in examining data from mobile operators. BPS uses thirteen indicators in conducting quality assurance, where the inspection uses several different methods, such as setting a threshold, checking data completeness, and checking the form of data distribution. Exploratory Data Analysis is carried out to determine whether the data meets the requirements for further analysis. We conducted this research on a mobile network operator data for June - July 2020 as the basis for MPD analysis in 2021. Based on the inspection during this period, BPS can cooperate with this cellular operator to conduct data analysis in 2021. However, the operator must repeat the calculation of the required matrix as quality assurance every month.