ARGUS/OVERWATCH
NAMES · WHO THE NEWS NAMED · 10-18 TO 10-31

Jantar Mantar

person Name of specific instruments in India. Wikidata

Named in 5212 articles by 76 newsrooms, in 408 stories, on 14 of the last 14 days, first on 10-18.

Rank
0.1268
by the company kept · 1 is the top name
Articles
5212
9518 mentions
Linked to
1550
names in 2+ shared articles
This week
0.29×
against last week
In Congress's circle
73%
of that hub's 30 closest names are also linked here
Bridging
0.0058
between groups · 1 is the top bridge

Most exclusively tied top 30 · ties that pass the test on the right, most exclusive first

The names that appear with Jantar Mantar in a large share of their own coverage: this name's own company. Exclusivity runs from 0, two names that meet no more than chance would have them, to 1, two names that only ever appear together. A name that is in everything, a president, is exclusive to no one, so the big names fall away here and the names that belong to this one stay. Only ties that pass the strength test count, so a pair seen twice is not a tie. Places are context and have no page.

NameKindExclusivityTogether
Cockroach Janta Partyorg0.7221411
Delhi Policeorg0.7171868
CJPorg0.712800
Gyanesh Kumarperson0.6751496
Saurav Dasperson0.6491377
Delhigpe0.6242542
Connaught Placeorg0.609458
Abhijeet Dipkeperson0.588921
CECorg0.571653
Central Delhigpe0.568222
New Delhigpe0.539911
Special Intensive Revisionorg0.531410
Narendra Modiperson0.503649
Tricolourgpe0.47978
Election Commissionorg0.463351
SIRorg0.463350
Neha Boraperson0.446308
Delhi High Courtorg0.437189
ACPorg0.43585
Bharatiya Janata Partyorg0.416233
Union Homeorg0.411143
Hyde Parkloc0.40916
Scrollorg0.40293
Gurugramorg0.40278
Aam Aadmi Partyorg0.489
Oppositionorg0.396137
Rahul Gandhiperson0.386426
Capitalgpe0.38473
ANIorg0.37366
Aadhaarorg0.37242

Most strongly tied the ties chance explains least

How sure the record is that two names appear together for a reason. Strength is a standard statistical test (log-likelihood): above 10.8, chance would produce the pairing less than once in a thousand times. It grows with the number of articles, so the biggest names come first here; the list on the left is where this name's own company shows.

NameStrengthTogether
CJP208672800
Delhi160872542
Delhi Police149931868
Cockroach Janta Party121121411
Gyanesh Kumar114331496
Saurav Das100721377
Abhijeet Dipke6275921
New Delhi5499911
CEC4527653
Narendra Modi3760649
Connaught Place3753458
Special Intensive Revision2781410
Election Commission2007351
SIR2001350
Central Delhi1874222

Articles per day

  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27
  • 28
  • 29
  • 30
  • 31

The hubs' circles share of each hub's 30 closest names that are also linked to this one

The eight most-named figures each have an inner circle: the thirty names most tied to them. This is how much of each circle this name is also tied to. A count of shared company in the coverage, nothing more.

Shared with Congress: AFP, ANI, Aadhaar, Aam Aadmi Party, Abhijeet Dipke, Amit Shah, Andhra Pradesh, Apple, Assembly, BJP, BNS, Bandra than Goa, Bench, Bharatiya Janata Party, Bollywood, CBI, CCTV, CEC, CJP, Cabinet.

As the sources put it relations a model read from the stories' sources · shown when two or more stories say it

The words are the sources' words, read out by a model; nothing here reads intent. A relation from one story alone is not shown.

  • 2 stories
    Saurav Das entered barricaded area and waved the national flag Jantar Mantar CJP leaders Saurav Das and Ashutosh Ranka released after seven days in custody

The year 9941 articles on 42 days, 2026-07-17 to 2026-10-31 · weekly bars · 14 months on record

Left is a year ago, right is this week. Older days come from headlines only, so their counts run lower than live days.

Named together, month by month · by surprise, among that month's articles · names with a page

2026-09 ECI 9 · Rahul Gandhi 9
2026-08 CJP 2 · Rahul Gandhi 2 · Congress 2
2026-07 Uttarakhand 2

Stories that name Jantar Mantar newest edition first · 40 shown

Names are found by a language model trained to spot them, joined to one spelling per person, and counted per article. The relations are the model's reading of the sources. Everything else is a count of who was written about together; it says nothing about why.