Abstract

This study examines the casualties occurring in the Israeli-Palestinian conflict spanning from October 2000 to September 2023. Using R software for comprehensive data processing and visualization, the research aims to uncover patterns in deaths related to the conflict. Through meticulous data cleaning, organizing, and employing various visualization methods, this investigation reveals trends in conflict-related fatalities. By utilizing techniques such as time series analysis and forecasting, the study identifies significant surges in casualties during specific periods. These findings, supported by multiple visual representations and statistical assessments, provide insights into critical phases characterized by heightened violence. The study’s revelations offer a deeper comprehension of the human cost within the Israeli-Palestinian conflict, presenting nuanced perspectives on periods marked by increased turmoil throughout the examined time frame.

Introduction

The Israel-Palestine conflict, characterized by a contentious struggle for land, identity, and sovereignty, traces its origins to the late 19th century. The rise of Zionism, advocating for a Jewish homeland, led to increased Jewish migration to Palestine within the Ottoman Empire. In 1917, the Balfour Declaration conveyed British support for establishing a Jewish national home in the region. Tensions escalated significantly in 1947 with the United Nations proposing a partition plan for Palestine, envisioning separate Jewish and Arab states. While Jewish leaders accepted the plan, Arab leaders rejected it, sparking conflict following Israel’s declaration of independence in 1948. Subsequent armistice agreements in 1949 defined Israel’s borders but left the West Bank under Jordanian control and the Gaza Strip under Egyptian control, exacerbating the Palestinian refugee crisis. The aftermath of the Six-Day War in 1967 saw Israel’s occupation of the West Bank, Gaza Strip, East Jerusalem, and the Golan Heights. This occupation facilitated Israeli settlements in these areas, contributing to ongoing tensions and conflicts. Despite peace efforts such as the Oslo Accords in 1993, aiming to address complex issues like land, security, and the status of Jerusalem, obstacles persisted. Contentions remained, particularly regarding Israeli settlements and conflicting claims to Jerusalem. From the 2000s onward, peace negotiations encountered obstacles amid intermittent violence and conflict. Disputes concerning borders, settlements, Palestinian refugees’ rights, and Jerusalem’s status hindered substantial progress towards a lasting resolution. This entrenched conflict is deeply rooted in historical, religious, and political complexities, presenting intricate challenges that defy straightforward solutions despite numerous international peace initiatives. Achieving a sustainable and fair resolution remains a formidable task amidst enduring hostilities and deeply entrenched grievances on both sides.

Motivation

This project strives to offer a factually grounded comprehension of the human toll within the Israeli-Palestinian conflict. Through thorough scrutiny of fatalities between 2000 and 2023 using advanced data analysis tools, it aims to reveal underlying patterns, pivotal periods, and shifts in violence. Unraveling these subtleties holds potential to aid policymakers, scholars, and humanitarian entities in formulating more well-informed approaches toward conflict resolution, promoting peace, and lessening the dire impact on civilians. Ultimately, the project seeks to harness insights derived from data analysis to contribute to raising awareness, nurturing empathy, and facilitating informed decision-making concerning this prolonged and intricate conflict.

Objectives

To examine the gender, nationality, and age demographics of individuals who have died.
To assess the weaponry employed during the conflict.
To evaluate the geographical distribution of fatalities across districts or regions.
To investigate the potential presence of seasonal patterns in the data.
To forecast forthcoming trends based on the analysis.

Data Preparation

To analyse the given data it was important to understand the data set and all the variables associated with it. The data set had total 11124 values. The variables of interest were gender, age, citizenship, weapons involved event locations and event dates. Vismiss plot was used to check for missing values. 4.7% of the data had missing values. To deal with it columns ammunitions and took_part_in_the_ hostilities were discarded from the study. The rows of remaining missing values were discarded. Effective methodology helped in removing 457 values. This helped in preventing the major loss of data and use maximum out of it. As age is a categorical variable it was converted to age groups. The dates also being a categorical variable it was converted into a few desired formats of MM-YYYY and YYYY for effective visualizations. The main challenge in the data set was the event locations. More than 80 cities were re-arranged into their respective districts/regions. For time-series analysis the dates were again re-arranged into ascending order and it was found to have some missing values. To deal with them Kalman filter was used. Kalman filter is valuable due to its ability to combine information from past and present measurements while considering uncertainties, resulting in more accurate predictions and estimates compared to raw or noisy data. Its adaptive nature and capability to handle noisy data make it widely applicable across various fields for improving the quality of estimates and predictions.

Results and Discussion

Between 2000 and 2009, the Israeli-Palestinian conflict witnessed several major incidents and conflicts:

Second Intifada (2000-2005): The Second Intifada, characterized by widespread Palestinian protests, violence, and suicide bombings, continued into the early 2000s. This period saw a surge in violence, resulting in a significant number of casualties on both sides.

Operation Defensive Shield (2002): In response to a wave of suicide bombings and attacks against Israeli civilians, Israel launched a large-scale military operation known as Defensive Shield in the West Bank. Israeli forces conducted extensive military operations, reoccupying several Palestinian cities and conducting targeted operations against militants.

Construction of the West Bank Barrier: During this period, Israel began the construction of the West Bank Barrier (referred to by Israelis as a security fence and by Palestinians as the separation wall). The barrier’s construction caused considerable controversy and heightened tensions due to its impact on Palestinian communities, land confiscation, and restrictions on movement.

Gaza Disengagement (2005): Israel unilaterally withdrew its settlers and military forces from the Gaza Strip in 2005, evacuating around 8,000 settlers from the area. Despite this disengagement, the conflict in Gaza continued due to ongoing border disputes, rocket attacks from Gaza into Israel, and Israeli military responses.

Gaza Wars (2008-2009): In late 2008, following a series of rocket attacks from Gaza, Israel launched Operation Cast Lead in December 2008. This offensive aimed at halting rocket fire and targeting Hamas militants in Gaza. The conflict resulted in significant casualties, infrastructure damage, and international condemnation of both sides for civilian casualties.

These conflicts and incidents between 2002 and 2009 contributed to a significant loss of life, infrastructure damage, and heightened tensions between Israelis and Palestinians, further complicating efforts towards peace and reconciliation in the region.

Between 2010-2013, direct talks failed, and Abbas pursued UN statehood recognition. The Gaza conflict “Operation Pillar of Defense” erupted in 2012, targeting Hamas. Secretary Kerry’s peace initiatives in 2013 aimed to restart negotiations but faced obstacles. Disputes over settlements and security persisted, stalling progress. These years were marked by intermittent violence and diplomatic endeavors, yet a comprehensive peace agreement remained elusive.

Compared to some other years within the Israeli-Palestinian conflict, the number of killings and casualties during the period from 2010 to 2013 was generally lower. While there were instances of violence, conflicts, and military operations during these years, the overall fatalities and casualties during this specific time frame were not as high as in certain other years, such as 2014, which witnessed a significant escalation and higher numbers of casualties.

In 2014, the Israeli-Palestinian conflict witnessed a significant surge in violence, notably with Israel’s military operation, Operation Protective Edge, launched in response to intensified rocket attacks from Gaza. Lasting 50 days, the conflict resulted in over 2,100 Palestinian deaths, primarily civilians, and numerous casualties on the Israeli side. The operation drew global attention and condemnation due to the scale of civilian casualties, including women and children, and extensive infrastructure damage in Gaza. The conflict eventually ceased following a mediated ceasefire, marking 2014 as a particularly deadly and impactful year within the Israeli-Palestinian conflict.

Between 2014-2022, Gaza saw intermittent conflicts in 2018 and 2021, with notable hostilities. US recognized Jerusalem as Israel’s capital in 2017. Israel forged normalization deals with some Arab nations in 2020. Settlement expansions persisted. While casualties occurred, they didn’t match 2014’s intensity. Peace efforts stalled due to disputes over borders, security, and settlements. The conflict’s persistent cycles of tension, occasional violence, and diplomatic shifts marked these years.

In the Israeli-Palestinian conflict, Jerusalem, particularly East Jerusalem, stands as a center of cultural and religious significance, triggering sporadic violence due to disputes over sovereignty and access to religious sites. Although Jerusalem has witnessed periodic clashes and protests, fatalities specifically linked to the conflict have been relatively fewer compared to regions like Gaza.

The Gaza Strip, notably during conflicts like Operation Protective Edge in 2014, confrontations in 2018, and escalations in 2021, has faced substantial casualties, primarily among Palestinians. Airstrikes, ground operations, and artillery fire resulted in significant loss of life in Gaza.

In southern Israel, communities bordering Gaza have encountered rocket attacks, causing occasional casualties among Israeli civilians and security forces. Although fewer in number compared to Gaza, these attacks have resulted in fatalities and injuries among Israelis.

While Jerusalem experiences tensions and occasional outbreaks of violence, casualties in the city due to the conflict have been comparatively lower, with Gaza bearing the brunt of casualties during intensified confrontations between Israelis and Palestinians.

Graphical Insights

You can access the number of deaths of the citizens by hovering your cursor to the respective bar of each bar chart and respective section of pie chart.

We can clearly see that the number of males is significantly higher (87.24%) than that of females (12.76%).

The people aged between 19-36 (58.48%) have been majorly involved in the conflict followed by the people aged less than 19 (24.15%). Becomingly 82.663% of the youth is involved in the conflict. This also adds to the fact that there are only 17.42% of the people who die after the age of 36 resulting to the major loss of youth.

Approx 90.79% deaths are observed of Palestinian nationality while only 9.21% approx belonged to Israel.

The 90.79% killings were done by Israeli nationals.

More than 55.13% of deaths happened in years between 2000-2010 but the count peaked in 2014 contributing to 21.49%.

Use of gun fire, shelling and explosives is observed highest from the word cloud.

Heat Maps

You can access the number of deaths of the citizens of both the countries by hovering your cursor on the white colored district/region borders in the heat maps.

The Gaza Strip overall has clearly faced a lot of casualties followed by Jerusalem and its surrounding areas.

Jerusalem and Southern District i.e HaDarom are the most affected areas by the casuilties followed by surrounding areas.

note: Golan District is not showed in this map. It does not affect the analysis.

Time Series

Time series analysis examines sequential data collected over regular intervals, identifying trends, patterns, and seasonality. It involves statistical methods to forecast future values based on historical data requiring trend analysis and prediction over time.

A time series plot with a regression line indicates the overall trend in sequential data. It represents the relationship between time and the variable of interest, showing if the variable increases, decreases, or remains constant over time, aiding in trend identification and prediction.

## 
##  Augmented Dickey-Fuller Test
## 
## data:  mts
## Dickey-Fuller = -6.3336, Lag order = 6, p-value = 0.01
## alternative hypothesis: stationary

The Augmented Dickey-Fuller (ADF) test assesses stationarity in time series data by testing for the presence of a unit root. Stationarity, indicating constant statistical properties over time, is vital for reliable modeling. The test examines a null hypothesis of a unit root’s presence, implying non-stationarity. Requirements include ensuring stationarity or transformability into a stationary series, appropriate lag order selection to capture data dependencies, and interpreting the test statistic against critical values. A rejected null hypothesis suggests stationarity, aiding in selecting suitable time series models like ARIMA and enhancing accurate forecasting.

A p-value of 0.01 in the context of hypothesis testing, such as in an Augmented Dickey-Fuller (ADF) test for unit root presence in a time series, suggests strong evidence against the null hypothesis. This result suggests that the series might exhibit stationary behavior, enabling more reliable modeling and forecasting in time series analysis.

## Series: mts 
## ARIMA(0,0,1) with non-zero mean 
## 
## Coefficients:
##          ma1     mean
##       0.3773  39.9151
## s.e.  0.0575   9.4919
## 
## sigma^2 = 13231:  log likelihood = -1700.36
## AIC=3406.73   AICc=3406.82   BIC=3417.59

ARIMA (AutoRegressive Integrated Moving Average) is a widely used time series forecasting method that models a series’ future values based on its own past values, integrating autoregression, differencing, and moving average components to capture patterns and trends.

An ARIMA(0,0,1) model with a non-zero mean signifies a time series model that includes a moving average (MA) component with order 1 and no autoregressive (AR) or differencing components. The non-zero mean suggests a constant term (μ) added to the model, indicating that the series has an average value different from zero. This model implies that the current value of the series depends on the error term from the previous period and the mean value, with no reliance on its own lagged values or differences for prediction.

In simple terms, The ARIMA(0,0,1) model with a non-zero mean suggests that recent changes or fluctuations in casualty incidents might influence future occurrences. Additionally, it implies a steady average level of casualties, independent of past incidents, existing throughout the entire period examined. This model assumes the future number of incidents is influenced by recent changes and a consistent average level of incidents over time.

ACF (AutoCorrelation Function) measures the correlation between a time series and its lagged values, showing how each point relates to past observations. PACF (Partial AutoCorrelation Function) displays direct correlations, removing indirect relationships, aiding in identifying significant lags for forecasting.

The ACF and PACF plots justify our ARIMA(0,0,1) model.

Forecasting in time series predicts future values based on historical data patterns, trends, and behavior. It utilizes statistical models to estimate and project forthcoming outcomes, aiding in making informed decisions. In our use case we have forecasted (indicated by a blue line and gray confidence interval) the possibility of casualties for next one year i.e 2024.

Conclusion, Shortcomings and Future Scope

The Israeli-Palestinian conflict involves territorial disputes over land, governance, and self-determination, marked by historical, political, and religious tensions between Israelis and Palestinians in the region. It becomes important to keep death count of the civilians different from that of security forces and terror groups to clearly differentiate with collateral damage but we can strongly consider the reason of casualties due to terrorism because of Hamas’s political and terror structure.

We could clearly examine the number of deaths depending on geopolitical dynamics which led to skirmishes observed to be peaked in 2014. The involvement of youth was also observed and maximum of them died before reaching age 36. Heat maps clearly showed Gaza strip and Jerusalem being the most affected areas due to religious and political disagreements.

Time series analysis showed that terrorism holds stationarity, stating that there is a trend and it can be best described by a first-order moving average model. Considering the ongoing war between both the countries, forecasting for becomes insignificant for available data set.

But forecasting can help in real time analysis of the data. In the future scope other forecasting methods such as Holt Winters forecasting, prophet models or similar machine learning models can be applied for better analysis. Moreover, the data from social media posts and happenings can effectively help in predicting the terror activities.